3

Phthalates and Male Reproductive-Tract Development

Phthalates are ubiquitous environmental chemicals that are anti-androgenic. They are found in a wide variety of consumer products, including toys, cosmetics, pharmaceuticals, and building and construction materials. Human exposure to phthalates has been well documented and occurs following ingestion, dermal exposure, or inhalation (Hauser and Calafat 2005; Lioy et al. 2015). Because of concerns about the toxicity of phthalates, the use of certain phthalates in children’s toys and child care articles has been regulated in the United States.¹ The European Union has also regulated the use of certain phthalates in toys, food-packaging materials, and cosmetics (EU 2004, 2005a,b, 2007). The National Health and Nutrition Examination Survey (NHANES) has documented widespread exposure to multiple phthalates in the general population (CDC 2009, 2015). An examination of temporal trends in phthalate exposure between 2001 and 2010 found reductions in the concentrations of some urinary phthalate metabolites and increases in the metabolite concentrations of replacement phthalates (Zota et al. 2014). Phthalates cross the placenta (Saillenfait et al. 1998; Fennell et al. 2004), and multiple phthalates have been measured in human and animal amniotic fluid (Silva et al. 2004; Calafat et al. 2006; Wittassek et al. 2009; Huang et al. 2016). In the rat, alterations in male reproductive-tract development after in utero exposure are the most sensitive health outcomes resulting from exposure to phthalates (NRC 2008; CHAP 2014). In rats, the anti-androgenic phthalates are those with ester side chains containing 4-10 carbon atoms, and some phthalates (e.g., dimethyl and diethyl phthalate) are not anti-androgenic or reproductive toxicants in the male rat (Gray et al. 2000; Furr et al. 2014).

Diester phthalates are initially hydrolyzed to their monoester, which undergoes subsequent glucuronidation and urinary excretion (see Figure 3-1). Other phthalate monoester metabolites can undergo additional oxidation of the alkyl side chains resulting in more complex metabolic profiles (Latini 2005; Calafat et al. 2006). For example, di(2-ethylhexyl)phthalate (DEHP) is metabolized to mono-2-ethylhexyl phthalate (MEHP), which undergoes additional oxidative side chain metabolism. Some representative phthalates and their oxidative metabolites are provided in Table 3-1. Biomonitoring efforts rely on the measurement of metabolite concentrations in urine (Samandar et al. 2009; Johns et al. 2015).

Male reproductive outcomes found in animal studies from in utero exposure to phthalates has been referred to as “phthalate syndrome” and include decreased anogenital distance (AGD), infertility, decreased sperm count, cryptorchidism (undescended testes), hypospadias (malformation of the penis in which the urethra does not open at the tip of the organ), and other reproductive-tract malformations (Gray et al. 2000; Foster et al. 2001; Fisher et al. 2003; NRC 2008). A hypothesized syndrome in the human (“testicular dysgenesis syndrome”) shares some of the same end points as the rat phthalate syndrome (Skakkebaek 2002; NRC 2008; Wohlfahrt-Veje et al. 2009); the etiology of the proposed human syndrome is unknown, however, and may or may not involve exposure to phthalates.

Phthalate male reproductive toxicity was one of the case examples the committee explored at a workshop held on February 3, 2016, which was designed to assist the committee with selecting the topics for its systematic reviews (see Appendix B for the workshop agenda and topics). Positive feedback was received from the participants at the meeting that there is an adequate data set to perform systematic reviews of the animal and the human evidence and to explore dose-response relationships on the effects of phthalates on male reproductive-tract development.

___________________

¹ Consumer Product Safety Improvement Act of 2008, Title II § 108 (a)(b) (H.R. 4040).

TABLE 3-1 Parent Phthalate and Oxidative Metabol ites Found in Urine Following Exposure

The committee focused its review on end points relevant to the anti-androgenic activity of phthalates, including fetal testosterone concentration, AGD, and hypospadias. A mechanistic link between decreased fetal testosterone levels and AGD and hypospadias is well established in animal models (e.g., Wilson et al. 2008; Scott et al. 2009). In rats, AGD is a well-known marker of androgen activity during the male programming window. Although the cause of hypospadias in humans can be multifactorial, mutations reducing androgen activity cause hypospadias in humans (van der Zanden et al. 2012). An association between hypospadias and reduced AGD has been observed in humans (Hsieh et al. 2012; Jain and Singal 2013; Thankamony et al. 2014), which suggests that human AGD is also dependent on androgen activity during the human male programming window—that is, the period during gestation when the male reproductive tract is programmed so that it will differentiate and grow normally (Ban et al. 2008; Hsieh et al. 2012; Dean and Sharpe 2013). The male programming window in the rat is gestation days 16-18, which corresponds to gestation days 14-16 in the mouse and approximately gestation weeks 8-14 in the human (Welsh et al. 2008).

Consideration was given to including cryptorchidism as an end point, but the committee decided against it for several reasons. Mechanisms for phthalate-induced cryptorchidism involve not only reduced fetal testis testosterone production but also reductions in fetal testis insulin-like 3 (INSL-3) production (Howdeshell et al. 2015). Rats exposed to phthalates have similar sensitivity to decreased fetal testosterone and AGD just as they do for decreased INSL-3 (Gray et al. 2016). In addition, cryptorchidism is a less sensitive end point compared to reductions in AGD (Saillenfait et al. 2008; Kim et al. 2010). Few human studies that examined the relationship between phthalate exposure and cryptorchidism were available to compare with animal data, which was also an important comparison to address the committee’s statement of task. Because the committee’s objective was to use the results of animal and human systematic reviews to evaluate the coherence between effects and dose-response relationships, the committee judged that including cryptorchidism in the analysis would not provide additional value to the project.

Two systematic reviews were conducted to answer the question what is the effect of in utero exposure to phthalates on AGD, hypospadias, or testosterone concentrations in males? One systematic review focused on animal studies and the other on human studies. This chapter first presents the methods that were used to conduct the two reviews. Then results of the reviews are presented together, along with mechanistic and other relevant information, to draw hazard conclusions.

SYSTEMATIC REVIEW METHODS

Protocols for the conduct of the systematic reviews were developed and peer reviewed. The PECO (Population, Exposure, Comparator, and Outcome) statements for the systematic reviews of the animal and the human studies are presented in Boxes 3-1 and 3-2, and the protocols used to conduct the systematic reviews are provided in Appendix C (Section C-1) and Appendix D (Section D-1), respectively. The protocols were based on the method developed by the National Toxicology Program’s Office of Health Assessment and Translation (OHAT) for conducting systematic reviews (hereto referred to as the OHAT method) (NTP 2015). A summary of the methods is briefly described below. The protocols were peer reviewed in accordance with standard report review practices of the National Academies of Sciences, Engineering, and Medicine. Most of the peer reviewers of the protocols were also peer reviewers of this report to ensure that the original protocols were followed and that any revisions or updates have been appropriately documented and justified. See the Acknowledgments for the list of peer reviewers.

Literature Searches and Screening

Scientific literature databases were searched for relevant studies on the effects of phthalates on male reproductive-tract development. A librarian, with specific training and expertise in performing searches for systematic reviews, developed and conducted the searches. A search for relevant existing systematic reviews was performed first, to avoid duplicating any recent work or work in progress. PubMed was

searched for systematic reviews published in 2013 or later, and the systematic-review protocol registries PROSPERO and CAMARADES were searched on August 3, 2016, for relevant protocols. Searches to support the systematic reviews were performed by the librarian in PubMed, Embase, and Toxline on August 15, 2016. The search strategies for animal and human publications are presented in the respective protocols (see Appendix C, Section C-1b, and Appendix D, Section D-1b).

References were screened at the title and abstract level and at the full-text level by the same two people using DistillerSR (https://www.evidencepartners.com). The screening criteria used are specified in the protocols (see Appendix C, Section C-1c, and Appendix D, Section D-1c). At the title and abstract screening level, if there was disagreement between the reviewers or an abstract was not available, the reference was passed on to the full-text screening level for further review. At the full-text level, disagreements about whether to include a reference were discussed by the two reviewers to reach agreement; if consensus could not be reached, a third team member was consulted to resolve the differences.

Data Extraction

Data from the included studies were entered into the Health Assessment Workspace Collaborative (HAWC), a Web-based interface application for warehousing data and creating visualizations (https://hawcproject.org). See Appendix C (Section C-1d) and Appendix D (Section D-1d) for data extraction elements for animal and human studies, respectively. One person entered data and a second person verified the entries. All data entered into HAWC are available at the following links: https://hawcproject.org/assessment/351/ (for the animal assessment) and https://hawcproject.org/assessment/350/ (for the human assessment).

Risk of Bias and Study Quality Evaluations

Risk of bias is related to the internal validity of a study and reflects study design characteristics that can introduce a systematic error (or deviation from the true effect) that might affect the magnitude and even the direction of the apparent effect. Internal validity or risk of bias was assessed for individual studies using a tool developed for the OHAT method that outlines an approach to evaluating risk of bias for experimental animal and human studies (NTP 2015). The risk of bias criteria were customized from the basic OHAT method and described in the protocol for addressing the specific research question for this review (e.g., methods for measuring AGD and fetal testosterone) (see Appendix C, Section C-1e, and Appendix D, Section D-1e). Key risk of bias elements in animal studies included reliability of the outcome measure, blinding of researchers to treatment groups, and the issue of whether investigators controlled for litter effects in their experimental design or statistical approaches. Key risk of bias elements in human epidemiologic studies included confounding, exposure characterization, and outcome assessment (including blinding of outcome assessors). Two committee members independently assessed each study and answered all applicable risk of bias questions following prespecified criteria detailed in the study protocol. One individual from each pair then reconciled any discrepancies with input from the second committee member. Any members who were the study author of a publication under review recused themselves from the evaluation of their study.

Data Analysis and Evidence Synthesis

For each outcome, the body of evidence was synthesized qualitatively and, where appropriate, a meta-analysis was performed. If a meta-analysis was performed, summaries of main characteristics for each included study was compiled and reviewed by two team members to determine comparability between studies, identify data transformations necessary to ensure comparability, and determine whether heterogeneity was a concern. The main characteristics considered across all eligible animal studies include the following:

• Experimental design (e.g., acute, chronic, multigenerational);
• Animal model used (e.g., species, strain, genetic background);
• Age of animals (e.g., at start of treatment, mating, and/or pregnancy status);
• Developmental stage of animals at treatment and outcome assessment;
• Dose levels, frequency of treatment, timing, duration, and exposure route;
• Health outcome(s) reported and their measurement;
• Type of data (e.g., continuous or dichotomous), statistics presented in the original publication; and
• Variation in degree of risk of bias at individual study level.

Uses of meta-analyses and meta-regression of experimental animal studies is provided in Box 3-3, and the methods used for performing meta-analyses, meta-regression, and benchmark dose estimation are summarized in Box 3-4.

The main characteristics considered across all eligible human studies include the following:

• Study design (e.g., cross-sectional, cohort);
• Details on how participants were classified into exposure groups (e.g., quartiles of exposure);
• Details on source of exposure data (e.g., questionnaire, area monitoring, biomonitoring);
• Measurement of biomonitoring data specific to phthalate exposure for each exposure group;
• Health outcome(s) reported;
• Conditioning variables in the analysis (e.g., variables considered confounders);
• Type of data (e.g., continuous or dichotomous), statistics presented in paper; and
• Variation in degree of risk of bias at individual study level.

Confidence Rating and Level of Evidence Conclusions

The quality of evidence for each outcome was evaluated using a grading system based on a modification of the GRADE system for rating the confidence in the body of evidence (Guyatt et al. 2011; Rooney et al. 2014). The process for rating the body of evidence as high, moderate, low, or very low was guided by the OHAT method (see Figure 3-2). In brief, studies on a particular outcome were initially grouped by key study design features, and each grouping of studies was given an initial confidence rating by those features. Several factors were then considered to determine whether the initial rating should be downgraded or upgraded. Factors that decrease confidence in results and lead to downgrading are risk of bias, unexplained inconsistency in results, indirectness or lack of applicability, imprecision, and publication bias. Factors that increase confidence in results and can upgrade a rating are these: a large magnitude of effect; evidence of a dose-response relationship; consistency across study designs, populations, animal models, or species; consideration of residual confounding; and other factors that increase confidence in the association or effect (e.g., rare outcomes). Confidence ratings were independently assessed by two committee members, and discrepancies were resolved by consensus and consultation with a third team member as needed. After a final confidence rating is determined, the rating is translated into a level of evidence using the scheme presented in Figure 3-3.

Integration of Evidence and Drawing Hazard Identification Conclusions

The committee used guidance from OHAT to draw hazard identification conclusions (NTP 2015). The procedure involves integrating the levels of evidence ratings for the human and animal data and considering them within the context of mechanistic information. The five possible hazard conclusions are (1) known, (2) presumed, (3) suspected, (4) not classifiable, or (5) not identified to be a hazard to humans. If either the animal or the human evidence stream has been described as having inadequate evidence, conclusions are drawn on the basis of a single evidence base. The hazard identification scheme is presented in Figure 3-4.

RESULTS

Literature Search and Screening Results

A search for existing systematic reviews on phthalate exposure and male reproductive-tract development in animals or humans found one publication in PubMed (Kay et al. 2014), but it was a literature review article and not a systematic review. No relevant protocols for ongoing systematic reviews were found in PROSPERO or CAMARADES.

A search of electronic databases for relevant publications to address the animal systematic review PECO statement found 1,527 unique citations (see Appendix C, Section C-2). A total of 311 publications met the criteria for full-text review, and 64 of them met the inclusion criteria for data extraction. A review of the reference lists of the 64 included studies identified an additional 16 publications that were potentially relevant. Those publications underwent the same screening process as did the publications found through database searches, and six publications met the inclusion criteria for data extraction (see Figure 3-5 for an illustration of the screening process and the exclusion criteria used at the full text screening level). Thus, animal data were extracted from 70 publications (see Box 3-5).

A search of electronic databases for human studies found 594 unique citations (see Appendix D, Section D-2, for a breakdown by database). A total of 27 publications met the criteria for full-text review, and 13 of them met the inclusion criteria for data extraction. A review of the reference lists of the 13 human studies identified an additional six publications that were potentially relevant. Those publications underwent the same screening process as the publications found through database searches, and three publications met the inclusion criteria (see Figure 3-6). A closer evaluation of the set of 16 included publications revealed that three of the publications (Adibi et al. 2015; Barrett et al. 2016; Martino-Andrade et al. 2016) involved “subanalyses” of a cohort by Swan et al. (2015) and one publication (Swan 2008) had expanded results from an earlier cohort by Swan et al. (2005) and had a larger sample size. To avoid double-counting data from the same cohort, the reports from Adibi et al. (2015), Barrett et al. (2016), and Swan et al. (2005) were excluded from data extraction. Martino-Andrade (2016) was retained because it provided additional information beyond Swan et al. (2015) on windows of exposure during the second and third trimester. Thus, data were extracted from 13 publications (see Box 3-6).

Health Effects Results

Effects of in utero phthalate exposure on male reproductive-tract development were evaluated separately for the human and the animal evidence. The outcomes examined were male AGD, fetal testosterone concentrations, and hypospadias incidence. Data were extracted from each of the studies and risk of bias assessments were performed. For the purposes of demonstrating the evaluations steps of rating the confidence in the bodies of evidence, performing qualitative and quantitative evidence syntheses, and drawing hazard identification conclusions, a decision was made to focus on a single phthalate. DEHP was selected for the example because it is a known anti-androgenic phthalate with widespread human exposures and was one of the congeners that had a robust set of human and animal studies. Results for other phthalates are also summarized later in this chapter.

Animal Health Effect Results on DEHP

Effects on AGD

Summary of the Evidence. There were 19 experimental animal studies that evaluated DEHP and AGD. Sixteen studies used the rat model and three studies used the mouse model (see Table 3-2). Phthalate exposure in all of the studies encompassed the entirety of the male programming window. Some of the data contained in the Wolfe and Layton (2005) study were from sire-only exposure and were not used in the committee’s analysis. Within some studies, AGD was measured at more than one postnatal age; in these instances, only data from the earliest postnatal age were used in the analysis because AGD may change during aging (McIntyre et al. 2001). Some studies presented AGD data in more than one manner (e.g., both corrected and uncorrected for body weight); in these instances, AGD data corrected for body weight were used.

Risk of Bias Considerations. Figure 3-7 shows the risk of bias evaluations of the studies used by the committee to assess DEHP effects on AGD. The primary factors of concern for animal studies are reliability of outcome measure, blinding of researchers to treatment groups, and control for litter effects. The majority of the studies did not adequately describe the method of AGD measurement and/or the reliability of the test methods used to measure AGD (e.g., use of micrometer caliper or reticule micrometer), and in most of the studies, blinding of the assessor was not reported. In addition, for the majority of studies the experimental design and/or statistical methods did not explicitly account for litter effects. Thus, most of the studies were rated as having a high risk of bias (or not reported) in these categories. The risk of bias assessment also considered when outcome assessments were performed (i.e., age), characterization of the test chemical, exposure methods, concealment of allocation to study groups, and information regarding attrition and data exclusion. Because data reporting of methods and results was often incomplete, numerous studies received a “not reported” rating for one or more of these secondary risk factors. There was no evidence of publication bias (see Appendix C, Section C-3).

Confidence in the Body of Evidence. The initial rating for the confidence in the animal studies was high because they involved controlled exposures, exposures occurred prior to outcome, outcomes were measured on individual animals, and a concurrent control comparison group was used (see Figure 3-2 for OHAT method for rating confidence). Confidence was downgraded because of the concern of significant risk of bias (described above under “Risk of Bias Considerations”) related to confidence in the reliability of outcome measure, blinding of investigators to the treatment groups, and control for litter effects. Confidence was upgraded because of a large magnitude of effect and because of evidence of a dose response.

TABLE 3-2 Summary of Animal Studies of DEHP and AGD

NOTE: BW, body weight; GD, gestation day; ND, not defined; PND, postnatal day.

A meta-analysis of studies of DEHP and AGD is presented later in this chapter (see “Meta-Analysis of DEHP and Reductions in AGD in Rats and Mice”). The results of the meta-analysis were subsequently factored into the following decisions regarding the committee’s confidence in the body of evidence

• Factors potentially decreasing confidence:
  • Unexplained inconsistency. No downgrade because most of the heterogeneity can be explained by dose, species, or strain. For instance, when separated by strain, and under a linear or linear-quadratic meta-regression in dose, there is no evidence of important heterogeneity in the rat data, with low values for I² that were not statistically significant. Under a linear or linear-quadratic meta-regression in dose, there is no evidence of important heterogeneity in the mouse data, with I² values of zero.²
  • Imprecision. The summary overall estimate, linear trend in log₁₀(dose), and linear trend in dose were all statistically significant in rats. Additionally, the statistical significance was robust under multiple sensitivity analyses. In contrast, the overall summary estimate for mice was not statistically significant; the linear trend in log₁₀(dose) and in dose were both statistically significant. This statistical significance was not robust under some sensitivity analyses, however. Therefore, the meta-analysis supports a downgrade in confidence based on imprecision in the mouse studies only. Because the mouse studies account for a small percentage of the overall body of evidence (three of 19 studies) the overall confidence in the body of evidence was not downgraded for imprecision.
• Factors potentially increasing confidence
  • Large magnitude of association or effect. In rats, the effects could be considered large and robust, with overall summary estimates having z-scores³ of ≥7.0. Moreover, these effect sizes were robust to multiple sensitivity analyses. Therefore, the meta-analysis supports an upgrade in confidence based on large magnitude of effect.

___________________

² I² describes the percentage of the variability in effect estimates that is due to heterogeneity rather than sampling error. The Cochrane Handbook provides the following guide to the interpretation of I² values: 0% to 40% (might not be important); 30% to 60% (may represent moderate heterogeneity); 50% to 90% (may represent substantial heterogeneity); and 75% to 100% (considerable heterogeneity) (Higgins and Green 2011).

³ Z scores are calculated by the ratio of the effect estimate Beta to the standard error, which can be calculated from the 95% CI. Specifically z = Beta*3.92/(CI, upper – CI, lower). Values of Beta and the CI are given in the Appendix C, Section C-5.

Table 3-3 presents the overall high confidence rating for the body of evidence on DEHP and AGD in rodents, and the details about how the ratings was determined is presented in Appendix C, Section C-4.

Level of Evidence in the Health Effect. A meta-analysis performed on studies on AGD and DEHP (see “Meta-Analysis of DEHP and Reductions in AGD in Rats and Mice” presented later in the chapter) found consistent evidence of a decrease in AGD after in utero exposure to DEHP in rats. Using the OHAT method (see Figure 3-3), a high confidence rating in the body of evidence (described above) and evidence of an effect result in a conclusion that there is a high level of evidence that fetal exposure to DEHP is associated with a reduction in AGD in male rats.

Alterations in Fetal Testosterone Concentrations

Summary of the Evidence. Twelve studies examining DEHP and fetal testosterone concentrations in animals were available (see Table 3-4), 11 in rats and one in mice. All the studies examined testosterone levels during fetal life. Unlike hypospadias and AGD analyses, studies measuring testosterone levels within fetal life but outside of the male programming window were included because fetal Leydig cell testosterone production sensitivity to phthalate exposure encompasses the entirety of fetal life when the testis is producing testosterone. The phthalate mechanism does not appear to involve an effect on pituitary-derived luteinizing hormone (Martinez-Arguelles et al. 2013); therefore, testosterone data were excluded from the analysis when the underlying fetal testis incubation method included agonism of the luteinizing hormone receptor.

TABLE 3-3 Profile of the Confidence in the Body of Evidence on DEHP and AGD in Animals

^a Moore et al. (2001); Borch et al. (2004); Jarfelt et al. (2005); Wolfe and Layton (2005); Andrade et al. (2006); Culty et al. (2008); Lin et al. (2008, 2009); Christiansen et al. (2009, 2010); Gray et al. (2009); Martino-Andrade et al. (2009); Vo et al. (2009); Li et al. (2013); Zhang et al. (2013); Jones et al. (2015).

^b Liu et al. (2008); Do et al. (2012); Pocar et al. (2012).

TABLE 3-4 Summary of Animal Studies of DEHP and Testosterone

* Depended on the supplier.

Risk of Bias Considerations.Figure 3-8 provides a summary of the risk of bias evaluation of the studies used by the committee to assess DEHP effects on testosterone. The primary factors of concern for animal studies are reliability of outcome measure, blinding of researchers to treatment groups, and control for litter effects. The majority of studies described the methods used to measure fetal testosterone and used measurement methods that the committee considered reliable (e.g., use of radioimmunoassay or enzyme-linked immunosorbent assay procedures). In the majority of studies, the experimental design and/or statistical methods accounted for litter effects. The risk of bias assessment also considered blinding of investigators to the treatment groups, but this factor was considered a secondary element that did not influence the committee’s confidence in the body of evidence. The committee also considered when outcome assessments were performed (i.e., age), characterization of the test chemical, exposure methods, concealment of allocation to study groups, and information regarding attrition and data exclusion. Because data reporting of methods and results was often incomplete, numerous studies received a “not reported” rating for one or more of these secondary factors. There was no evidence of publication bias (see Appendix C, Section C-3).

Confidence in the Body of Evidence. The initial rating for the confidence in the animal studies was high because they involved controlled exposures, exposures occurred prior to outcome, outcomes were measured on individual animals, and a concurrent control comparison group was used (see Figure 3-2 for OHAT method for rating confidence). Confidence in the body of evidence was not downgraded for any factors, but was upgraded because of evidence of a large magnitude of effect and a dose response. A meta-analysis of studies on DEHP and fetal testosterone is presented later in this chapter (see “Meta-Analysis of DEHP and Alterations in Fetal Testosterone in Rats”), and informed decisions regarding the committee’s confidence in the body of evidence, including

• Factors potentially decreasing confidence
  • Unexplained inconsistency. No downgrade was warranted because some of the heterogeneity is explained by dose, but there was substantial residual variance. The size of the effect is large enough, however, so that concerns about inconsistency were not serious from the point of view of causal inference.
  • Imprecision. No downgrade was warranted because the overall summary estimate, linear trend in log₁₀(dose), and linear trend in dose were all statistically significant. Additionally, the statistical significance was robust under multiple sensitivity analyses.

• Factors potentially increasing confidence
  • Large magnitude of association or effect. Especially at higher doses, the effects on fetal testosterone could be considered large and robust, with overall summary estimates having z-scores of ≥7.0 and an overall summary estimate indicating >50% decreases. Moreover, these effect sizes were robust to multiple sensitivity analyses. Therefore, the meta-analysis supports an upgrade in confidence based on the large magnitude of effect.
  • Dose response. Upgraded because of strong evidence of dose response through meta-regression with statistically significant linear trends in either log₁₀(dose) or dose. Moreover, these results were robust to multiple sensitivity analyses.

Table 3-5 presents the overall confidence ratings for the body of evidence on DEHP and fetal testosterone in rodents, and the details about how the ratings were determined are presented in Appendix C, Section C-4. There is high confidence in the body of evidence from experimental studies in animals.

Level of Evidence in the Health Effect. A meta-analysis performed on studies on DEHP and fetal testosterone (see “Meta-Analysis of DEHP and Alterations in Fetal Testosterone in Rats” presented later in the chapter) found consistent evidence of a decrease in fetal testes testosterone after in utero exposure to DEHP in rats. Using the OHAT method (see Figure 3-3), a high confidence rating in the body of evidence and evidence of an effect result in a conclusion that there is a high level of evidence that fetal exposure to DEHP is associated with a reduction in fetal testosterone in rats.

Hypospadias

Summary of the Evidence. Nine studies of DEHP and hypospadias were available. Of these studies, eight used the rat model and one used the mouse model (see Table 3-6). The exposure paradigm for all studies included the entirety of the male programming window. Animal hypospadias data were collected on a litter and/or an individual animal basis, and both methods were considered in the analysis. The hypospadias detection method for all rat studies was visual inspection of the phallus during postnatal life. Unlike the rat studies, the mouse study used a unique assessment methodology (urethral casting) and examined the phallus during fetal life (gestation day 19) (Liu et al. 2008).

TABLE 3-5 Profile of the Confidence in the Body of Evidence on DEHP and Fetal Testosterone Concentrations in Animals

^a Do et al. (2012).

^b Borch et al. (2004, 2006); Culty et al. (2008); Howdeshell et al. (2008); Lin et al. (2008); Martino-Andrade et al. (2009); Vo et al. (2009); Hannas et al. (2011b); Klinefelter et al. (2012); Saillenfait et al. (2013a); Furr et al. (2014).

Risk of Bias Considerations. Figure 3-9 provides a summary of the risk of bias evaluation of the studies used by the committee to assess DEHP effects on hypospadias. The primary factors of concern for animal studies are reliability of outcome measure, blinding of researchers to treatment groups, and control for litter effects. The majority of studies did not adequately describe the method by which offspring were evaluated for hypospadias, and in 44% of the studies, blinding of the assessor was not reported. Most of the studies controlled for litter effects in the experimental design and/or statistical methods. The assessment also considered when outcome assessments were performed (i.e., age), characterization of the test chemical, exposure methods, concealment of allocation to study groups, and information regarding attrition and data exclusion. Because data reporting of methods and results was often incomplete, numerous studies received a “not reported” rating for one or more of these secondary risk of bias evaluations. There was no evidence of publication bias (see Appendix C, Section C-3).

Confidence in the Body of Evidence. The initial rating for the confidence in the animal studies was high because they involved controlled exposures, exposures occurred prior to outcome, outcomes were measured on individual animals, and a concurrent control comparison group was used (see Figure 3-2 for OHAT method for rating confidence). Confidence was downgraded because of the concern of significant risk of bias related to confidence in the outcome measure and blinding of investigators to the treatment groups. Confidence was also downgraded because of the concern of significant inconsistency in responses seen across studies. For example, for litters affected,⁴ only one study was available in Sprague-Dawley rats (Saillenfait et al. 2009), which reported increased incidences of hypospadias (>30%) at the tested doses of 500 and 625 mg/kg-day, whereas among the three studies in Wistar rats, only one study reported effects, with a low incidence (9%) and only at a single intermediate dose (300 mg/kg-day). More studies reported effects as percent of animals affected, with all three studies in Sprague-Dawley rats reporting increased incidences (>10%, up to 100%), and two studies in Wistar rats reporting small increases in incidence (up to 5%) at an intermediate dose of 300 mg/kg-day (from a range up to 900 mg/kg-day).

TABLE 3-6 Summary of Animal Studies of DEHP and Hypospadias

NOTE: BW, body weight; GD, gestation day; ND, not defined; PND, postnatal day.

___________________

⁴ Because of possible litter correlation, effects reported as percent of litters affected are preferred over effects reported as percent of animals affected.

Confidence in the body of evidence was upgraded because the background control incidence of hypospadias was reported as zero across all studies, so any positive finding was considered treatment related (i.e., rare outcome). Because hypospadias represents a dichotomous measure (present/absent) the presence of numerous studies reporting no incidence prevented the committee from completing a meta-analysis of the animal hypospadias data.

Table 3-7 presents the confidence ratings for the body of evidence on DEHP and hypospadias in rodents, and the details about how the rating were determined are presented in Appendix C, Section C-4. Overall there is moderate confidence in the body of evidence for hypospadias in animals.

Level of Evidence in the Health Effect. As described above, there is evidence of increased incidence of hypospadias in rats after fetal exposure to DEHP. Using the OHAT method (see Figure 3-3), a moderate confidence rating in the body of evidence and evidence of an effect result in a conclusion that there is a moderate level of evidence that fetal exposure to DEHP is associated with an increased incidence of hypospadias in male rats. The data suggest that Sprague-Dawley rats might be more sensitive to DEHP than Wistar rats.

TABLE 3-7 Profile of the Confidence in the Body of Evidence on DEHP and Hypospadias in Animals

^a Liu et al. (2008).

^b Jarfelt et al. (2005); Andrade et al. (2006); Christiansen et al. (2009, 2010); Gray et al. (2009); Saillenfait et al. (2009); Vo et al. (2009); Li et al. (2013).

Meta-Analyses of Animal Data

Meta-Analysis of DEHP and Reductions in AGD in Rats and Mice

The animal database for AGD and DEHP was judged to be amenable for meta-analysis. A summary of the analysis is provided below, and supporting details are presented in Appendix C, Section C-5. (Meta-analyses of studies on AGD and other phthalates are provided in Appendix C, Section C-6.) The following exclusions and groupings of studies were made to focus the analysis:

• Rat and mouse data were analyzed separately, due to known anticipated species differences in sensitivity (Johnson et al. 2012). Additionally, the rat data were subjected to a subgroup analysis by strain because of anticipated differential sensitivity across strains (Wilson et al. 2007).
• Studies in which exposures did not cover the entire male programming window (GD 16-18 in rats) were excluded.
• In many cases multiple AGD measures were reported for the same experiment. The measures selected for meta-analysis were in the following order of priority:
  • For studies that reported AGD at multiple time points in the same animals, the earliest postnatal time point was used.
  • For studies that reported AGD in multiple units, the order of preference was: AGD in mm/cube root of body weight, AGD in mm/body weight, and AGD in mm.

In all, 13 of the 16 rat studies and all three of the mouse studies were included in the analysis; three of the rat studies were excluded because they were missing group size values (Borch et al. 2004; Vo et al. 2009; Jones et al. 2015). Effect sizes were calculated as the log ratio of the mean difference between the treatment group and the concurrent control, multiplied by 100 (y = 100 × ln [mean of treated group ÷ mean of control group]). For small changes, this is approximately equal to the percent change, but the resulting confidence interval is more symmetric and closer to normal (Hedges et al. 1999; Lajeunesse 2011). This normalization allows for treatment groups to be compared across studies and experiments. When normalized in this way, however, treatment groups within a study are correlated. Therefore, in one of the sensitivity analyses, effects were estimated using only the highest treatment group from each study. Additional sensitivity analyses were performed by sequentially excluding each study (all treatment groups for that study) (see Appendix C, Table C5-2).

Both an overall effect of any treatment and coefficients of meta-regressions were estimated. For meta-regressions, three models were used: a linear model in y = a + b*log₁₀(dose) to test for a dose-response trend; a linear model y = b*dose; and a linear-quadratic model y = b*dose + c*dose² to model the dose-response shape. In the linear and linear-quadratic model, the intercept was omitted because the effect measures were already normalized relative to control levels. Additionally, for these models, the coefficients were rescaled in terms of the change per 100 mg/kg-day (e.g., y = b*[dose/100] + c*[dose/100]²) for ease of interpretation. In all cases, random effect models were used, as described in the protocol. All analyses utilized random effects models, as implemented in the R “metafor” package. Sensitivity analyses included leaving one study out at a time and using only the highest dose group in each study (see Appendix C, Table C5-2). Benchmark dose estimates were calculated for an effect size of 5% (BMD₅; see Appendix C, Tables C5-3 and C5-5). The BMD₅ was calculated using the linear or linear-quadratic model, with the model selection based on the lowest AICc (Akaieke information criterion corrected for small sample size). The BMD₅ was calculated only for the “fixed effect”—that is, the estimated mean response across studies.

The results in rats for AGD are as follows:

• Statistically significant overall effect of a reduction in AGD (–3.96 [95% confidence interval (CI)]: –5.07, –285) and linear trends in log₁₀(dose) (–1.97 [95% CI: –2.98, –0.96]) and dose (–1.55 [95% CI: –1.86, –1.24]). The overall effect was robust to leaving out individual studies.
• Under the linear-quadratic model, there was low heterogeneity (23%, p = 0.12), with a BMD₅ estimated to be 270 mg/kg-day (95% CI: 180, 420).
• When analysis was restricted to the highest dose group, there was a larger overall effect, larger linear trend in log₁₀(dose), consistent linear trend in dose, and consistent BMD₅ estimates.
• In subgroup analyses, there were statistically significant overall effects and linear trends in log₁₀(dose) and dose for Sprague-Dawley and Wistar rats separately, with reduced heterogeneity. Sprague-Dawley rats appeared somewhat less sensitive than Wister rats, with smaller overall effect sizes, smaller trend in log₁₀(dose), and larger benchmark dose estimates. Specifically, a BMD₅ for Sprague-Dawley rats was estimated to be 290 mg/kg-day (95% CI: 170, >1,000), whereas the BMD₅ for Wistar rats was estimated to be 150 mg/kg-day (95% CI: 100, 280).

The results of linear-quadratic meta-regression, the model with the lowest AICc, are shown in Figure 3-10.

The results in mice for DEHP and changes in AGD are as follows:

• No statistically significant overall effect, but statistically significant linear trends in log₁₀(dose) (–1.77 [95% CI: –2.71, –0.83]) and dose (–2.03 [95% CI: –3.51, –0.55]).
• Under the linear-quadratic model (–5.71 [95% CI: –7.15, –4.27]), there was low heterogeneity (0%, p = 0.19), with the BMD₅ estimated to be 110 mg/kg-day (95% CI: 90, 150).
• When analysis was restricted to the highest dose group, there remained no statistically significant overall effect, and there was no longer a statistically significant linear trend.
• Overall effect was no longer statistically significant when leaving out some individual studies during the sensitivity analyses.

The results for the overall effect estimate, which had the lowest AICc, are shown in Figure 3-11.

Overall there is consistent evidence of a decrease in AGD in male rats after fetal exposure to DEHP, with a modest dose-response gradient. After fitting a meta-regression linear-quadratic model, heterogeneity was low or not detectable. In rats, the effects were robust to sensitivity analyses (see Appendix C, Table C5-2), involving removal of individual studies and use of only the highest exposure group. In mice, the effect estimates were similar after removing individual studies or restricting to the highest dose group, but in some cases, they lacked statistical significance owing to larger confidence intervals (see Appendix C, Table C5-4). Sprague-Dawley rats are less sensitive than Wistar rats, with a BMD₅ of around 300 mg/kg-day compared to 150 mg/kg-day. Mice have a BMD₅ of 250-350 mg/kg-day, which is in the range of the Sprague-Dawley rat.

Meta-Analysis of DEHP and Alterations in Fetal Testosterone in Rats

The animal database for DEHP and alterations in fetal testosterone was amenable to meta-analysis. A summary is provided below, and supporting details are presented in Appendix C, Section C-5. The same meta-analysis approach that was used to evaluate AGD was also applied to studies in rats of DEHP and fetal testes testosterone (there was only one study in mice). The same exclusions/groups were made, with the additional consideration that effects reported at least 6 h after dosing in acute studies were preferred over effects reported at earlier times because an effect should be greater at later time points. In all,

7 of the 11 rat studies were ultimately included. The studies by Borch et al. (2004, 2006) and Vo et al. (2009) were excluded because they had missing group size values, and the study by Klinefelter et al. (2012) was excluded because testosterone measurements were taken after stimulation of the testes with luteinizing hormone. Benchmark dose estimates for effect size of 5% (BMD5) and 40% (BMD40) were calculated (see Appendix C, Table C5-8).⁵ The results are as follows:

• Statistically significant overall effect (–110.14 [95% CI: –136.73, –83.54]) and linear trends in log₁₀(dose) (–132.83 [95% CI: –171.03, –94.63]) and dose (–23.01 [95% CI: –26.24, –19.72]), with an overall effect that is large in magnitude (>50% change). The overall effect was robust to leaving out individual studies.

___________________

⁵ BMD₄₀s were calculated for this end point because previous studies have shown that reproductive-tract malformations were seen in male rats when fetal testosterone production was reduced by about 40% (Howdeshell et al. 2015; Gray et al. 2016).

• Under the linear-quadratic model (–34.23 [95% CI: –47.02, –21.44]), there remains a substantial, statistically significant heterogeneity (I² >95%, p <0.001), with the BMD₅ estimated to be 15 mg/kg-day (95% CI: 11, 24). The BMD₄₀ was found to be 160 mg/kg-day (95% CI: 120, 240).
• When analysis was restricted to the highest dose group, a larger overall effect, a larger linear trend in log₁₀(dose), a consistent linear trend in dose, and consistent benchmark dose estimates were found.
• In subgroup analyses, there were statistically significant overall effects and linear trends in log₁₀(dose) and dose for Sprague-Dawley and Wistar rats separately. Heterogeneity was reduced among Wistar rats (I² = 21%), but not among Sprague-Dawley rats (I² >95%).
• In subgroup analyses, there were statistically significant overall effects and linear trends in log₁₀(dose) and dose for Sprague-Dawley and Wistar rats separately, with reduced heterogeneity. Sprague-Dawley rats appeared to be slightly more sensitive than Wister rats, with slightly larger overall effect size and trend in log₁₀(dose) and slightly lower benchmark dose estimates. The BMD₅ for Sprague-Dawley rats was estimated to be 13 mg/kg-day (95% CI: 9, 23), whereas the BMD₅ for Wistar rats was estimated to be 23 mg/kg-day (95% CI: 21, 24). The corresponding BMD₄₀ estimates were 140 mg/kg-day (95% CI: 100, 230) for Sprague-Dawley rats and 230 mg/kg-day (95% CI: 210, 240) for Wistar rats.

The results of meta-regressions are shown in Figure 3-12 (linear-quadratic for Sprague-Dawley and linear for Wistar).

Overall there is consistent evidence of a decrease (>50% change) in fetal testes testosterone after DEHP treatment, with a strong dose-response gradient. Even after subgrouping by strain and meta-regression with dose, however, substantial heterogeneity remained in Sprague-Dawley rats. All three strains are outbred, so some of the residual heterogeneity may be due to genetic diversity. Nonetheless, the effects were robust to sensitivity analyses involving removal of individual studies and use of only the highest exposure group. Based on benchmark dose estimates, Sprague-Dawley rats are slightly more sensitive to these effects than Wistar rats (in contrast to the case with AGD).

Human-Health Effects Results on DEHP

Effects on AGD

Summary of the Evidence. The most robust data sets on DEHP were on AGD as measured by either AGD (ap [anopenile]) or AGD (as [anoscrotal]). The six epidemiologic studies that examined the relationship between biomarkers of DEHP exposure and AGD (ap or as) outcomes were all prospective cohort studies that enrolled pregnant mothers and their infants (see Table 3-8). A study by Suzuki et al. (2012) calculated and reported AGD index rather than AGD.

The cohort studies were performed in the United States (Swan 2008; Swan et al. 2015; Martino-Andrade et al. 2016); Scandinavia (Bornehag et al. 2015; Jensen et al. 2016); and Mexico (Bustamante-Montes et al. 2013). The studies varied in timing of when urinary phthalate metabolites were measured during pregnancy, age when infant AGD was measured, and the reporting on reliability of AGD measurements. Jensen et al. (2016) and Bustamante-Montes et al. (2013) measured urinary phthalate metabolites in the third trimester only, whereas Swan (2008) measured them throughout pregnancy but on average late in pregnancy. Bornehag et al. (2015) measured urinary phthalate metabolites during the first trimester of pregnancy.

The results of a fifth study from the same cohort were reported in two publications: Swan et al. (2015) used measurements in women who were less than 13 weeks pregnant, and Martino-Andrade et al. (2016) used measurements of second and third trimester urinary phthalate metabolites. AGD measurements were performed on infants up to 36 months of age in the studies by Swan (2008) and Bornehag et al. (2015), whereas measurements were taken in infants who were 3 months old or younger in the studies by Bustamante-Montes et al. (2013), Swan et al. (2015), Jensen et al. (2016), and Martino-Andrade et al. (2016). Reports of AGD measurement reliability varied across studies, as some included data on intra- and inter-rater reliability and others did not. (Appropriate methods involve standardized training for all examiners using calipers as the primary measurement instrument and continued repeat measurements on the same subject as well as by different trained examiners on the same subject throughout the study to

ensure low intra-rater and inter-rater variability.) No study was excluded on the basis of failing to report rater reliability, however, as long as measurement methods were appropriate and well described. All studies used state-of-the-art analytical chemistry methods to measure urinary phthalate metabolites, and they included collection of and adjustment for important potential confounding variables such as measures of infant body size and maternal demographic factors.

Risk of Bias Considerations. The risk of bias ratings for the individual studies are presented in Figure 3-13. The primary factors of concern for human studies are confounding, exposure characterization, and outcome assessment (including blinding of outcome assessors). The questions used to evaluate risk of bias in the individual studies are provided in Appendix D, Section D-1e. For this data set, the studies had either a low or a very low risk of bias in these domains. Specifically, the risk of bias assessment for urinary phthalate metabolite measurements (biomarkers of exposure) considered the reliability of the test methods (e.g., use of high performance liquid chromatography with tandem mass spectrometry) and whether the exposure biomarker was assessed in a relevant time-window for development of the outcome. It also considered whether there was a measure of urinary dilution that was accounted for in the analysis, such as urinary specific gravity or creatinine. The short half-life of DEHP (<24 h) (Koch et al. 2004) may contribute to exposure misclassification, an important issue in environmental epidemiologic studies. The gold standard would be multiple 24-h urine samples during the relevant sensitive window of exposure, which is difficult to obtain in human studies. Although one spot urine sample is not the best measure of long-term exposure over the relevant prenatal programming period, exposure misclassification would likely introduce random noise and bias toward the null.

Risk of bias evaluation of the outcome assessment considered the methods for determining the outcome, whether the outcome had been assessed consistently across all groups, and whether the outcome assessors had been blinded to the study groups or exposure levels prior to assessing the outcomes. Given the study designs for the epidemiologic studies that examined AGD, it was unlikely that examiners measuring AGD would know urinary phthalate levels at the time the AGD measurement was made.

Risk of bias assessment of the studies of DEHP and AGD also included assessment for important confounding variables such as age, race/ethnicity, weight/body size, and age at exam. Most studies measured multiple urinary phthalate metabolites in addition to DEHP metabolites. There can be a correlation among phthalate metabolites from different diesters. This may contribute to confounding by other metabolites. There was no evidence of publication bias (see Appendix D, Table D3-3).

Confidence in the Body of Evidence. The initial rating for the confidence in the human studies was moderate based on the following three criteria: exposures occurred prior to outcome, outcomes were measured on individuals, and a (control) comparison group was used (see Figure 3-2 for OHAT method for rating confidence). A meta-analysis of these data is presented later in this chapter, and it provides additional information concerning the confidence ratings. Specifically, meta-analysis supports the following:

• Factors potentially decreasing confidence
  • Unexplained inconsistency. No downgrade is warranted because the meta-analysis I² statistic was 0%. In some cases, larger values (up to 54%) were estimated in sensitivity analyses, but these are given less weight because they involved the use of less preferred outcome or exposure estimates, which are expected to introduce more heterogeneity.
  • Imprecision. The meta-analysis also supports that imprecision in the results is not a concern; the summary estimate has a 95% confidence interval of −6.49, –1.66, and the confidence intervals for the sensitivity analyses were similar. Therefore, the same causal conclusion would be reached based on either end of the confidence interval. As discussed in the GRADE framework (Guyatt et al. 2011), confidence intervals that would result in different conclusions depending on whether the upper or lower limit is used can result in a downgrade due to imprecision. Because the summary estimate and its statistical significance is robust to multiple sensitivity analyses, the meta-analysis would support the conclusion that imprecision is not a serious concern.

TABLE 3-8 Summary of Human Studies of DEHP and AGD

• Factors potentially increasing confidence
  • Large magnitude of association or effect. An upgrade is not warranted. Although the effect size observed for AGD of a 4% decrease per 10-fold increase in DEHP metabolite concentration can be considered relatively large—as this degree of change in AGD in experimental animal studies is associated with around a 40% decrease in fetal testosterone production—the smaller end of the confidence interval is an effect size of –1.66%, and in some of the sensitivity analyses, the smaller end of the effect size is <1%. Therefore, a small magnitude of effect cannot be ruled out with reasonable confidence.
  • Dose response: The effect estimates of AGD are estimates of slopes; thus, they are based on the assumption of a monotonic dose-response relationship between exposure and effect. One study reported dose-response information independent of slope estimates, and it was not informative due to wide confidence intervals. Therefore, the meta-analysis would not support an upgrade in the confidence conclusion based on evidence of a dose-response gradient.

There were no changes in the confidence rating for the human evidence after considering factors that could increase or decrease confidence. Table 3-9 presents an evidence profile of the findings on DEHP and AGD in humans, and additional details about how the moderate rating was determined is presented in Appendix D, Section D-3.

Level of Evidence in the Health Effect. The results show a consistent pattern of findings that higher maternal urinary concentrations of DEHP metabolites during pregnancy (during the prenatal male genital programming window) are associated with a smaller AGD in male infants compared to infants whose mothers had lower DEHP exposures during pregnancy. Consistent reductions in AGD were found across multiple studies; the small amount of heterogeneity observed may be due to sample size differences, AGD measurement variability, urinary metabolite concentration variability, and the potential for residual confounding. A meta-analysis (presented later in this chapter) found consistent evidence of a decrease in AGD being associated with increasing urinary concentrations of the sum of DEHP metabolites. Using the OHAT method (see Figure 3-3), a moderate confidence rating in the body of evidence (described above) and evidence of an effect result in a conclusion that there is a moderate level of evidence that fetal exposure to DEHP is associated with a reduction in AGD.

TABLE 3-9 Profile of the Confidence in the Body of Evidence on DEHP and AGD in Humans

^a Swan et al. (2008); Bustamante-Montes et al. (2013); Bornehag et al. (2015); Swan et al. (2015); Jensen et al. (2016); Martino-Andrade et al. (2016).

Testosterone Concentrations Measured During Gestation or at Delivery

For fetal testosterone assessment, only three human studies met the criteria for inclusion of having measurements of urinary concentrations of DEHP metabolites, either sumDEHP metabolites or individual DEHP metabolites. Another study was excluded from consideration because the authors measured only phthalate monoester metabolites (MEHP) in maternal blood. Because of concern with external contamination in matrices such as blood (Calafat et al. 2015), studies that measured only monoester phthalate metabolites in blood were excluded. Of the three studies, one study measured testosterone concentrations in amniotic fluid (Jensen et al. 2015), one measured testosterone in cord blood (Lin et al. 2011), and the other measured maternal serum testosterone concentrations during pregnancy (Sathyanarayana et al. 2014). The Jensen et al. (2015) study utilized a large biobank of amniotic fluid samples collected in Denmark between 1980 and 1996 to study the cross-sectional association of amniotic fluid concentrations of testosterone with oxidative metabolites of DEHP (and DINP). Within this biobank, they conducted a nested case-control study with 270 cases of cryptorchidism, 75 cases of hypospadias, and 300 male controls. Among the controls there was no association of amniotic fluid concentrations of mono-(2-ethyl-5carboxypentyl) phthalate (MECPP) with testosterone concentrations. The Lin et al. (2011) study measured maternal urinary concentrations of DEHP metabolites and found no association with cord serum concentrations of testosterone among male newborns. The Sathyanarayana et al. (2014) study was not further considered because it is not known if maternal serum concentrations directly reflect fetal testosterone production; therefore, it would not be possible to make inferences about the association of urinary DEHP metabolites with fetal testis testosterone production.

Based on the disparate matrices used to estimate fetal testis testosterone production (amniotic fluid or cord blood), the differences in timing of measurement of testosterone (during pregnancy or at delivery), and the dearth of studies, the committee determined that the data were inadequate to draw any conclusions. The committee also recognized that human studies on fetal testosterone production in relation to phthalate exposure are logistically very difficult to conduct, and it is not possible to directly determine fetal testis testosterone production. Although the Jensen et al. (2015) study used a design that most closely approximates assessing fetal testis testosterone production during pregnancy, its interpretation was hindered by uncertainty regarding relevance of amniotic fluid levels of testosterone to fetal testis testosterone production: namely, the pharmacodynamics of testosterone levels within amniotic fluid. Furthermore, uncertainty regarding the pharmacokinetics of fetal phthalate metabolism limited the interpretation of MECPP concentrations in amniotic fluid.

Hypospadias

For hypospadias, two human studies (Chevrier et al. 2012; Jensen et al. 2015) were available on measures of DEHP metabolites. Jensen et al. (2015) conducted a case-control study on hypospadias that was nested within a large biobank of amniotic fluid samples collected in Denmark between 1980 and 1996. They measured amniotic fluid MECPP concentrations among 75 cases of hypospadias and 300 controls. There was no association of MECPP with odds of hypospadias. The Chevrier et al. (2012) study, a nested case-control study with 21 cases of hypospadias, did not find an association between hypospadias and urinary concentrations of DEHP metabolites. These two studies used disparate matrices to measure DEHP metabolites (amniotic fluid and urine) and were very small in size, limiting their power. Given that these were the only two studies, the committee determined that the data were inadequate to draw any conclusions or conduct a meta-analysis.

Meta-Analysis of Human Data on AGD and DEHP

The epidemiologic database on ADG and DEHP was judged to be amenable for meta-analysis. Five cohorts contributed data to the analysis (Swan 2008; Bustamante-Montes et al. 2013; Bornehag et al. 2015; Swan et al. 2015; Jensen et al. 2016). The preferred measures for each study were:

• Outcome: AGD (as) is preferred over AGD (ap) because it is a more reliable measurement.
• Time of exposure measurement: The first trimester is preferred over the second trimester, which is preferred over the third trimester, because the male programming window is in the first trimester.
• Exposure metric: Sum of DEHP metabolites is preferred over MEHP, which is preferred over any of the other DEHP metabolites, because the sum better reflects the parent compound exposure.

The primary study data, using the preferred DEHP exposure biomarkers and outcome measure for each study, are presented in Table 3-10, and the result of the analysis is presented in Figure 3-14. For the studies by Bustamonte-Montes et al. (2013) and Swan (2008), the CIs were estimated using the reported p-value, assuming a normal distribution. For other studies, confidence intervals were included in the published manuscript.

TABLE 3-10 Summary of Human Data Used in Meta-Analysis of DEHP and AGD

NOTE: AGD (ap), distance between the anus and base of the phallus; AGD (as), distance between the anus and scrotum.

Slopes (beta coefficients) are reported in units of change in mm/log₁₀ change in urinary concentrations of DEHP metabolites. Two factors a priori may affect comparability across studies. First, there are baseline differences in AGD (as) across different studies due to such demographic factors as age at measurement of AGD, which is affected by weight and body size. For instance, the mean AGD (as) reported by Bustamante-Montes et al. (2013) was 12.4 mm, whereas the mean AGD (as) reported by Bornehag et al. (2015) was 41.4 mm. Second, AGD (as) is shorter than is AGD (ap). For instance, in the study by Jensen et al. (2016), mean AGD (as) was 36.9 mm, whereas mean AGD (ap) was 70.2 mm. Therefore, the same change in distance may reflect different percentage change in AGD across studies in end points. To standardize effect sizes across studies, each reported beta coefficient was divided by the mean value of the reported outcome measure prior to conducting the meta-analysis. The result is that each beta coefficient is standardized to a percent change in AGD per log₁₀ change in urinary DEHP metabolite concentrations.

All analyses utilized random effects models, as implemented in the R metafor package. Sensitivity analyses included leaving one study out at a time; using alternative exposure and outcome measures for each study, one at a time; and restricting analyses to use of the same exposure measure (the sum of DEHP metabolites or MEHP) and/or the same outcome measure AGD (as) or AGD (ap). Figure 3-15 shows the sensitivity analyses that were performed by leaving one study out at a time.

In the primary analysis, five studies, with beta coefficients standardized to a percent change per log₁₀ change in DEHP exposure, were analyzed using a random effects model. A statistically significant summary estimate of –4.07 (95% CI: –6.49, –1.66; [p = 0.001]) was found for the change in AGD per log₁₀ increase in DEHP exposure. There was no significant heterogeneity, with an estimated I² value of 0% (Q statistic was not statistically significant). Two studies (Swan 2008; Swan et al. 2015) accounted for over 60% of the weight in the summary estimate.

Leaving one study out at a time, the summary estimates ranged from –4.35 to –3.59. The summary estimate remained statistically significant in all cases, with p-values ranging from 0.001 to 0.019. There was no observed heterogeneity in any of these cases (I² value of 0%). After the Swan studies, the next largest weight in the summary estimate was from Jensen et al. (2016).

Sensitivity analyses were further performed using alternative effect estimates for each study (see Appendix D, Table D4-1). The summary estimates ranged from –4.78 to –1.51. In 11 of the 42 alternative analyses, the summary estimates were no longer statistically significant (summary estimates range from –1.51 to –2.69), with p-values ranging from 0.050 to 0.41. All of the nonstatistically significant alternative analyses involved replacing the Swan et al. (2015) results with results from Martino-Andrade et al. (2016) using second trimester or third trimester DEHP metabolite measurements. Each of these analyses also led to greater heterogeneity (I² up to 54%, though none were statistically significant).

Finally, eight additional sensitivity analyses were conducted restricting the included results to more homogeneous exposure and/or outcome measures (e.g., using only the sum DEHP metabolite estimates) (see Appendix D, Table D4-1). The resulting summary estimates ranged from –4.2 to –2.0, all of which were statistically significantly different from 0. Additionally, there was no observed heterogeneity in any of these cases (I² = 0).

Overall, there is consistent evidence of a decrease in AGD being associated with increasing urinary concentrations of the sum of DEHP metabolites and of magnitude around 4% for each log₁₀ increase in DEHP concentrations. There was no evidence of heterogeneity in the primary analysis, and this result was robust to removing individual studies. The result was also robust to 50 additional sensitivity analyses that used alternative effect size estimates. In about 80% of these sensitivity analyses, the summary estimate remained statistically significant. Moreover, the eight sensitivity analyses involving stricter criteria for homogeneous exposure and outcome measures had summary measures that were statistically significant with no observed heterogeneity. The majority of the weight in the preferred summary estimate, however, is from two studies from different cohorts (i.e., independent study populations) with the same first author

(Swan 2008; Swan et al. 2015). Dropping both of these studies would result in a summary estimate that is consistent with all analyses (negative indicating a reduction in AGD with a log₁₀ increase in DEHP) but was no longer statistically significant (not shown, –2.48 [95% CI: –6.42, 1.45], I² = 0). Overall, however, greater weight is given to the primary analysis because it includes all the available studies that met the prespecified inclusion criteria and because it reflects the preferred measures of outcome and exposure.

EVIDENCE INTEGRATION FOR AGD

Evidence synthesis for AGD was conducted in a three-part process. First, the confidence ratings for the human and animal studies were translated into conclusions about level of evidence of health effects using the procedure outlined by OHAT (performed earlier in this chapter). Second, an initial hazard identification conclusion was reached by integrating the conclusion about level of evidence for the human and the animal evidence streams. Third, the degree of support from mechanistic data was considered and discussed in reaching final hazard identification conclusions for AGD.

Initial Hazard Conclusion for AGD

As described in earlier sections, the level of evidence for fetal exposure to DEHP being associated with reductions in AGD was high for the animal evidence and moderate for the human evidence. Using the OHAT hazard identification scheme (see Figure 3-4), an initial hazard conclusion is reached that DEHP is presumed to be a reproductive hazard to humans. The human and animal bodies of evidence present a consistent pattern of findings that prenatal exposure to DEHP is associated with reduced AGD.

Consideration of Mechanistic Data on AGD

Mechanistic data available from the rat support that the following steps are involved in phthalate reproductive toxicity: (1) metabolism of the phthalate diester to the monoester; (2) decreased expression of genes that regulate cholesterol metabolism and steroidogenesis in fetal Leydig cells; (3) decreased production of fetal testis testosterone; and (4) reduced expansion of the perineum resulting in a change in AGD and altered urethral closure resulting in hypospadias (Howdeshell et al. 2015; see Figure 3-16). Several of these elements will be discussed in greater detail below.

Phthalate Metabolism and Pharmacokinetics

A critical first step in phthalate reproductive toxicity involves the hydrolysis of the diester phthalate to the more toxic monoester metabolite. Subsequent metabolic steps include formation of the MEHP-glucuronide by UDP-glucuronyl transferase and the creation of oxidized metabolites formed by cytochrome P450 4A that are further oxidized by alcohol or aldehyde dehydrogenases (Albro and Lavenhar 1989).

One of the more significant species differences in metabolism relates to the rate at which MEHP is formed by lipase. For example, lipase activity seen in mouse liver homogenates was markedly higher than that observed in marmosets (1,339 ± 261 pmol/g versus 62 ± 11 pmol/g) (Ito et al. 2005). Notable species differences in Vmax, Km, and Vmax/Km ratio were also seen between mice, rats, and marmosets, suggesting that species differences in lipase activity may result from different enzyme affinities and different expression levels of the enzyme (Ito et al. 2005). Follow-up studies have confirmed that humans, like marmosets, have lower hepatic lipase activity when compared with mice although the inter-individual variability is quite large in people (Ito et al. 2014). Other species differences in enzyme activity are seen with several of the other enzymes involved in DEHP metabolism; however, the extent of the difference is not as large as that seen with lipase (Ito et al. 2005).

The reproductive toxicity of a structurally related phthalate metabolite, monobutylphthalate (MBP), has been evaluated in pregnant marmosets. Fetal exposure of marmosets to MBP at 500 mg/kg-day during gestation weeks 7-15 did not affect plasma testosterone levels at birth or hypospadias or other changes in reproductive-tract development (McKinnell et al. 2009). Several marmosets exposed to DEHP in utero did develop clusters of undifferentiated testicular germ cells of unknown biological significance, however (McKinnell et al. 2009).

Additional pharmacokinetic studies have been performed in rodents, marmosets, and people. For example, Kessler et al. (2004) found that peak blood concentrations of MEHP in rats was approximately three times higher (range: 1.3-7.5 µg/mL) than in marmosets when both species were exposed similarly to DEHP. In addition to differences in metabolism there may also be differences in the way conjugated DEHP metabolites are eliminated. For example, several free oxidized DEHP metabolites are observed in the plasma of rats but not of marmosets (which more quickly glucuronidate these metabolites) following DEHP exposure (Kurata et al. 2012). Kessler et al. (2012) and Koch et al. (2004) evaluated the pharmacokinetics of DEHP in adult human volunteers after they ingested deuterium-labelled DEHP. A striking observation in the human pharmacokinetic study was that peak concentrations (C_max) and area under the curve (AUC) for MEHP and DEHP in human serum are much greater than those reported for either rats or marmosets given comparable administered doses. Kessler et al. (2012, p. 289) concluded that “the MEHP blood burden at a given DEHP dose per kg body weight will be higher in humans than in the animals.” Similar studies have not been performed in pregnant women.

The pharmacokinetics of DEHP has also been investigated in chimeric mice transplanted with human hepatocytes, experiments that also supported development of a simplified physiologically based pharmacokinetic (PBPK) model for DEHP (Adachi et al. 2015). The PBPK model consists of gastrointestinal, liver, and central compartments. Resulting PBPK model predictions suggest that MEHP will be cleared from plasma in humans similarly to what is seen in mice, but fecal elimination of MEHP (and other oxidized metabolites) will have a higher rate in people (Adachi et al. 2015). An important caveat remains, however, that placental transfer of phthalates to the fetus is incompletely understood and described in this model.

Decreased Expression of Genes That Regulate Steroidogenesis

Studies have shown that in utero exposure of animals to DEHP can produce a reduced expression of proteins involved in steroidogenesis in the fetal testis. Affected proteins can include CYP11A1, CYP17A1, translocator protein (18-kDa), and STAR (Gray et al. 2000; Borch et al. 2006; Culty et al. 2008). Although changes in STAR and other enzymes are seen following in utero exposure, knowledge concerning the molecular initiating event involved in these reductions in enzyme activity remains unclear. This data gap is not addressed in current high-throughput assay systems (e.g., ToxCast) because the steroidogenic assay used in these programs often relies on a human adrenal cell line (Karmaus et al. 2016) and adrenal steroidogenesis in vivo is not affected by phthalate exposure via the same mechanism (i.e., decrease in steroidogenic pathway gene expression) (Thompson et al. 2005; Martinez-Arguelles et al. 2011).

Decreased Production of Fetal Testis Testosterone

The committee’s systematic review found a high level of evidence that in utero exposure to DEHP in rats is associated with a reduction in fetal testosterone levels. Sprague-Dawley rats appeared to be slightly more sensitive to this effect than Wister rats. The BMD₅s for Sprague-Dawley and Wistar rats were estimated to be, respectively, 13 (95% CI: 9, 23) and 23 (95% CI: 21, 25) mg/kg-day. In contrast, based on the committee’s BMD₅ estimates for AGD, Sprague-Dawley rats were approximately twofold less sensitive to DEHP-induced effects on AGD when compared with Wistar rats, suggesting possible strain differences in the quantitative relationship between decreases in fetal testosterone and changes in AGD.

Studies have shown that reproductive-tract malformations were found in male rats when fetal testosterone production was reduced by about 25-70% (Howdeshell et al. 2015). The association between decreases in fetal testosterone and changes in AGD in other species is less clear. For example, studies conducted in mice with a structurally related phthalate, di-n-butyl phthalate (DBP), have shown that reduced fetal testicular testosterone occurs in rats, but not in mice, following in utero exposure (Gaido et al. 2007; Johnson et al. 2012).

Biological Plausibility

The mechanistic data developed in vitro and in animal models provide evidence that the DEHP effects on AGD in humans identified by the committee’s systematic review are biologically plausible. Moreover, androgen-dependent development of the male reproductive tract and androgen-dependent AGD appear to be well conserved across mammalian species (including humans). Nevertheless, the mechanistic data were not sufficient to result in an upgrade in the committee’s final hazard identification for AGD (see Figure 3-4).

Final Hazard Conclusion on AGD

On the basis of the committee’s evidence integration of the animal and the human evidence on DEHP and effects on AGD and consideration of relevant mechanistic data, the committee concluded that DEHP is presumed to be a reproductive hazard to humans.

EVIDENCE INTEGRATION FOR FETAL TESTOSTERONE

The approach to integrating the animal and the human evidence on the effects of DEHP on fetal testosterone was the same as that used for AGD.

Initial Hazard Conclusion for Fetal Testosterone

As described in earlier sections, the level of evidence for fetal exposure to DEHP being associated with decreased fetal testosterone synthesis was high for evidence in rats and was inadequate for evidence

in humans. Using the OHAT hazard identification scheme (see Figure 3-4), an initial hazard conclusion was reached that DEHP is presumed to be a reproductive hazard to humans.

Consideration of Mechanistic Data

Decreased Testosterone Following DEHP Exposure

As mentioned earlier, mechanistic data available on the rat support the hypothesis that decreased production of fetal testis testosterone occurs in animals following fetal exposure to DEHP. The decrease in production in rats may be secondary to reduced expression of STAR and other proteins involved in fetal testis steroidogenesis (Gray et al. 2000; Borch et al. 2006; Culty et al. 2008). These in vivo studies are supported by some in vitro studies that demonstrated decreased testosterone production in cultured rat fetal testes exposed to MEHP but not the parent phthalate DEHP (Chauvigné et al. 2009). Other studies conducted with cultured rat testes failed to show an effect of MEHP at in vitro concentrations up to 10 μM (Stroheker et al. 2006). Likewise, these effects have not been observed in cultured human fetal testes treated with MEHP (Lambrot et al. 2009). Pharmacokinetic studies of DEHP and other phthalates suggest that differences in decreased fetal testes testosterone production reflect differential potency for testosterone inhibition rather than differences in tissue dosimetry (Clewell et al. 2010).

Another line of mechanistic data the committee considered was the result of xenograft studies performed in rodents for exploring differences in species sensitivity. In one of these experiments, fetal rat, mouse, and human testes were implanted in nude rats or mice exposed to DBP at 250 or 500 mg/kg-day for 1-3 days (Heger et al. 2012). Only rat xenografts exhibited statistically significant decreases of steroidogenic gene expression (including Star) and testosterone secretion (Heger et al. 2012). As with implanted mouse testes, human testes did not develop statistically significant phthalate-induced suppression of steroidogenic gene expression (including Star) following host exposure up to 500 mg/kg-day for two days (Heger et al. 2012). Similar results have been obtained in experiments that examined the responses of human fetal tissues (collected at gestational weeks 10-23) implanted in castrated immunodeficient mouse hosts for 6 weeks (Mitchell et al. 2012). Qualitatively, the data indicate that human testes appear to be less sensitive to the effects of phthalates than the rat testes. A limitation of those studies is that they rely on small numbers of human tissue samples. Consequently, the committee performed a literature search on February 9, 2017, and identified two studies that provide xenograft data regarding testosterone concentrations (Mitchell et al. 2012; Spade et al. 2014). Mean and standard errors were digitized and standard errors were converted to standard deviations. The effect measure is the log ratio of the mean between treated and control, times 100 (which for small values is close to the percent change). Random effects models were fit for overall effect. There were too few studies to do sensitivity analyses. The overall effect size was estimated to be –15.7 (95% CI: –51.8, 20.4), corresponding to a percent change of –14.5% (95% CI: –40.4, 22.6). There was no heterogeneity observed (I² = 0%). While a trend toward decreased serum testosterone was observed, it was not statistically significant. Due to the low precision of the estimate, however, the data are inadequate to conclude whether an effect may have occurred, since they are consistent with effect sizes ranging from a 40% decrease to a 23% increase in serum testosterone. Figure 3-17 presents the results of the meta-analysis.

Biological Plausibility

The mechanistic data developed in vitro and in animal models provides evidence that DEHP effects on testosterone in rats identified by the committee’s systematic review is biologically plausible. However, the mechanistic data were not sufficient to result in an upgrade in the committee’s final hazard identification for fetal testosterone.

Final Hazard Conclusions for Fetal Testosterone

On the basis of the committee’s evidence integration of the animal and the human evidence on DEHP and effects on fetal testosterone and consideration of relevant mechanistic data, the committee concluded that DEHP is presumed to be a reproductive hazard to humans.

EVIDENCE INTEGRATION FOR HYPOSPADIAS

The approach to integrating the animal and the human evidence on the effects of DEHP on fetal testosterone was the same as that used for AGD.

Initial Hazard Conclusion for Hypospadias

As described in earlier sections, the level of evidence for DEHP being associated with increased incidence of hypospadias was moderate for the animal evidence and inadequate for the human evidence. Using the OHAT hazard identification scheme (see Figure 3-4), an initial hazard conclusion is reached that DEHP is suspected to be a reproductive hazard to humans.

Consideration of Mechanistic Data

As mentioned earlier, mechanistic data available from the rat support the hypothesis that hypospadias and other phenotypic changes observed in “testicular dysgenesis syndrome” are dependent on reduction in testosterone production by the fetal Leydig cell as a proximate cause (Howdeshell et al. 2015). The results of the committee’s systematic review demonstrating that DEHP induced decreases in fetal testosterone following in utero exposure provide indirect evidence for this hypothesis.

The linkage between phthalate exposure, decreased testosterone, and phenotypic changes in other species remains uncertain. For example, testicular histopathology (multinucleated gonocytes) in fetal mice exposed to DBP at 500-1,500 mg/kg-day occur in the absence of a significant decrease in testicular testosterone (Gaido et al. 2007; Lehraiki et al. 2009). Presumably, the development of microscopic evidence of urethral changes in mice (Liu et al. 2008) following DEHP exposure would also occur in the absence of decreased fetal testis testosterone production. Understanding of DEHP’s effects on human fetal testosterone production and hypospadias is limited to in vitro studies and human-rodent xenograft studies that do not recapitulate the intact organism. As mentioned earlier in the discussion of the fetal testosterone data, human fetal testes cultured in vitro or implanted in rodent hosts in xenograft experiments often behave similarly to mouse testes and do not develop significant changes in fetal testosterone production following DEHP or MEHP exposure (Lambrot et al. 2009; Mitchell et al. 2012; Spade et al. 2014).

Biological Plausibility

The mechanistic data developed in vitro and in animal models provide additional evidence that the DEHP effects on testosterone and hypospadias in rats identified by the committee’s systematic review are biologically plausible. Mechanistic data on DEHP effects on human fetal testosterone production and hypospadias were largely lacking, however. When considered collectively, the mechanistic data were not sufficient to result in an upgrade in the committee’s final hazard identification for hypospadias.

Final Hazard Conclusions for Hypospadias

On the basis of the committee’s integration of the animal evidence and the human evidence on DEHP and fetal hypospadias and consideration of relevant mechanistic data, the committee concluded that DEHP is suspected to be a reproductive hazard to humans.

CONSIDERATION OF LOW-DOSE EFFECTS

As described above, DEHP is presumed to be a reproductive hazard to humans on the basis of evaluations of the evidence on AGD and fetal testosterone. The data on AGD in animals and humans were sufficiently robust to allow quantitative characterization of the dose-response relationships, whereas the fetal testosterone evidence was not. The human data provide moderate evidence of a relationship between DEHP exposure and decreases in AGD, with a magnitude decrease of 1.7-6.5% for each log₁₀ increase in exposure. When the experimental animal data were analyzed in the same manner (estimating the magnitude of change for each log₁₀ increase in exposure), the effect estimates are similar: 0-2% for SpragueDawley rats, 1-5% for Wistar rats, and 1-3% for mice. Thus, these estimates are by and large concordant; however, the dose ranges in which these estimates have been observed differ substantially between humans and rats.

A direct comparison of DEHP exposure in animal and human studies included in the systematic reviews was not possible because exposure is measured differently in the studies. The epidemiologic studies measured DEHP metabolites in maternal urine, and none performed dose reconstruction to estimate phthalate intake or to estimate levels of active DEHP metabolites in blood or fetal testis. In contrast, animal studies reported external (administered) doses and reported no metabolite measurements in urine or other specimens. Comparison of rat administered doses and predicted human DEHP intake was, therefore, based on other studies.

Estimates of human daily intake of DEHP are generally made using either pharmacokinetic models that predict intake from urinary concentrations of phthalate metabolites (“reverse toxicokinetics”) or by estimating the fraction of DEHP excreted in the urine within 24 h (Koch 2004; Lorber et al. 2010; Anderson et al. 2011; Kessler et al. 2012). Human intake estimates were available from studies that used urinary measurements reported for the general US population (Lorber et al. 2010), German subjects (Wittassek and Angerer 2008), and the Taiwanese population exposed to phthalates because of its illegal use in food

and beverages (Chang et al. 2017). Estimates of mean daily intake in adults in the US population range from approximately 0.0006-0.002 mg/kg-day (Lorber et al. 2010) to 0.011 mg/kg-day (Lorber and Calafat 2012). The study of German subjects estimated a median (maximal) daily intake of 0.003 (0.042) mg/kg-day (Wittassek and Angerer 2008). Chang et al. (2017) estimated median daily intakes of DEHP in the Taiwanese population to be about 0.004 mg/kg-day for men and 0.002 mg/kg-day for women (maximal intakes were greater than 0.008 mg/kg-day). Estimates of daily DEHP intake in all three reports were several thousand times lower than the BMD₅s of 15 and 150 mg/kg-day calculated by the committee (see Table 3-11).

Urinary MEHP concentrations in rats and humans were evaluated. Urinary MEHP concentration has been measured in rats treated with DEHP at 11 mg/kg-day (Calafat et al. 2006), a dose that is close to the lowest BMD₅ of 15 mg/kg-day calculated by the committee. Urinary MEHP concentration in pregnant rats 6 h after dosing averaged 1,626 ng/mL (Calafat et al. 2006). In humans, maternal urinary MEHP concentration was about 16.5 ng/mL (highest concentration reported for the 75th percentile group in the epidemiology studies in Table 3-11) or approximately 100 times lower than mean urinary concentrations measured in rats. Also, the 95th percentile for urinary MEHP concentrations in the general adult US population (38.9 ng/mL) is only 40 times lower than that of rats dosed at 11 mg/kg. Lorber et al. (2010) has suggested that the survey data used may underestimate US population exposure to DEHP because measurements are made in spot urine samples collected during the day, often from fasting participants. Spot urine samples have high temporal variability, especially in response to bolus dosing, so they might not provide an accurate reflection of the circulating metabolites.

A comparison of MEHP in amniotic fluid in rats dosed at 11 mg/kg-day, near the BMD₅ for decreased testosterone, and humans found that MEHP concentrations were similar. One study included in the human systematic review reported a median amniotic fluid MEHP concentration of 22.8 ng/mL (Huang et al. 2009). This concentration is about three times lower than the mean MEHP concentration (68 ng/mL) found in amniotic fluid of pregnant rats exposed at 11 mg/kg-day (Calafat et al. 2006). Slightly larger differences were found in comparisons with other human studies that were not included in the committee’s systematic review (Silva et al. 2004; Wittassek et al. 2009; Huang et al. 2016).

These comparisons, qualitatively similar to margin-of-exposure comparisons, produce strikingly different results depending on whether the comparisons are based on estimates of DEHP intakes or measurement of urinary or amniotic MEHP concentration. The finding that MEHP levels in amniotic fluid in the general US population are in the same range as in rats treated near the BMD₅ for decreased fetal testosterone suggests that the dose-response relationships that emerged from the meta-analyses of the animal and human data are generally concordant. Much less is known about MEHP levels in serum or testis in humans compared with rats. As noted earlier, a striking observation in the human pharmacokinetic study is that the peak concentrations and area under the curves for MEHP and DEHP in human serum are much greater than those reported for either rats or marmosets given comparable administered doses (Koch 2004; Kessler et al. 2012). On the basis of these pharmacokinetic differences in serum MEHP and DEHP concentrations, humans could be 2-100 times more sensitive than rats to DEHP (Koch 2004; Kessler et al. 2012).

Inferences about effect levels in humans and rats are uncertain for a number of reasons. In general, the relationship between intake of DEHP and urinary levels of it and its metabolites is fairly well described, but there is much less confidence in information about blood or fetal testis concentrations in rodents or humans and their relationships with intake levels. The maternal urinary and amniotic fluid measures in rats were from a single study (Calafat et al. 2006), and the study had several limitations. Only two rats per dose group were used in the study; however, five urinary samples were collected per rat and a correlation between administered dose and urinary or amniotic fluid MEHP was found. The authors noted that MEHP was primarily found in the urine as the glucuronide conjugate, which was not consistent with other studies that reported that rats excrete free MEHP. Amniotic fluid should reflect a more integrated exposure with less variability, but there are few human studies that measure phthalate metabolites in amniotic fluid and even fewer that measured them in rats. Even though pharmacokinetic differences between humans and rats or marmosets have been demonstrated, and may explain why humans might be more

sensitive to DEHP, little information on partitioning of DEHP or its metabolites to the fetal compartment in humans is available.

OTHER PHTHALATES

This section summarizes the committee’s evaluation of the other phthalates: BzBP, DBP, DEP, DIBP, DINP, and DPP. The same methods that were used to evaluate DEHP were used to evaluate these phthalates as well. Details about the risk of bias evaluations and how confidence in the body of evidence was rated are provided in Appendix C, Section C-4, for the animal studies and in Appendix D, Section D-3, for the human studies. The sections below will briefly discuss how the available data compare with DEHP and will provide initial hazard identification conclusions for AGD, fetal testosterone, and hypospadias.

Animal Health Effect Results for Other Phthalates

AGD

As with DEHP, phthalate exposure in all the animal studies of other phthalates encompassed the entirety of the male programming window, and AGD measurements taken at the earliest postnatal age were used in the analysis. For any meta-analyses performed, AGD data corrected for body weight were used whenever possible. The same risk of bias factors as with DEHP were considered, with the key factors being reliability of the outcome measure, blinding of investigators to the treatment groups, and control for litter effects. The risk of bias assessment of the animal studies also considered when outcome assessments were performed (i.e., age), characterization of the test chemical, exposure methods, concealment of allocation to study groups, and information regarding attrition and data exclusion. There was no evidence of publication bias for any of the other phthalates (see Appendix C, Section C-3).

The initial rating for the confidence in the evidence from animal studies was high because they had controlled exposures, exposures occurred prior to outcome, outcomes were measured on individual animals, and a concurrent control comparison group was used (see Figure 3-2 for OHAT method for rating confidence).

BzBP and AGD

Confidence in the Body of Evidence. There is high confidence in the body of evidence on developmental exposure to BzBP and effects on AGD in rodents. Six studies examining BzBP and AGD in rats were available (see Table 3-12). Five studies in rats found decreased AGD after developmental exposure to BzBP (Ashby et al. 1997; Nagao et al. 2000; Ema and Miyawaki 2002; Tyl et al. 2004; Aso et al. 2005). Confidence in the evidence was determined by considering factors that might upgrade or downgrade confidence (see Figure 3-2 for the factors). Confidence was downgraded because of risk of bias concerns; all studies had ratings of probably high risk or definitely high risk of bias in at least one of the key issues considered, and they had multiple risk of bias issues (see Figure 3-18). Confidence in the BzBP evidence was upgraded because three studies showed a relatively large magnitude of change (about 20-40%) in the same dose range, and most studies reflected a similar magnitude of response within the same dose range (see Appendix C, Figures C4-12 and C4-13).

Level of Evidence in the Health Effect. Using the OHAT method (see Figure 3-3), a high confidence rating in the body of evidence and evidence of an effect result in a conclusion that there is a high level of evidence that BzBP is associated with a decrease in AGD in male rats.

TABLE 3-11 Comparison of Human and Rat Intake and Internal Concentrations of DEHP

* For boys with shorter AGD.

NOTES: CI, confidence interval; GM, geometric mean; LOD, limit of detection; ND, not detected; NHANES, National Health and Nutrition Examination Survey.

TABLE 3-12 Studies of BzBP and AGD in Rats

NOTES: The earliest life stage evaluated is shown when multiple observation times were assessed. GD, gestation day; LOAEL, lowest-observed-adverse-effect level; NOAEL, no-observed-adverse-effect level; PND, postnatal day.

DBP and AGD

Confidence in the Body of Evidence. There is high confidence in the body of evidence on developmental exposure to DBP and effects on AGD in rodents. Twenty-two studies examining DBP and AGD in rodents were available, and multiple studies in rats found decreased AGD after developmental exposure to DBP (see Table 3-13). Confidence was downgraded because of risk of bias concerns; all studies had ratings of probably high risk or definitely high risk of bias in at least one of the key issues considered, and most of the studies had multiple risk of bias issues (see Figure 3-19). Confidence in the DBP evidence was upgraded because most studies reflected a similar magnitude of response within the same dose range (see Appendix C, Figures C4-12 and C4-13).

TABLE 3-13 Studies of DBP and AGD in Rats

NOTES: The earliest life stage evaluated is shown when multiple observation times were assessed. GD, gestation day; LOAEL, lowest-observed-adverse-effect level; NOAEL, no-observed-adverse-effect level; PND, postnatal day.

Level of Evidence in the Health Effect. Using the OHAT method (see Figure 3-3), a high confidence rating in the body of evidence and evidence of an effect result in a conclusion that there is a high level of evidence that DBP is associated with a decrease in AGD in male rats.

DINP and AGD

Confidence in the Body of Evidence. There is very low confidence in the body of evidence on developmental exposure to DINP and effects on AGD in rodents. Four studies examining DINP and AGD in rats were available (see Table 3-14). Only one study found decreased AGD after developmental exposure to DINP (Boberg et al. 2011). Confidence was downgraded because of risk of bias concerns. Two of the studies had a probably high risk of bias rating in two key areas (whether researchers were blinded to the treatment groups or how outcomes were assessed), and one of them had a probably high risk of bias rating for not controlling for litter effects. Another study also had a no reporting about whether the researchers were blinded. Confidence in the evidence was also downgraded because of unexplained inconsistency and imprecision (see Appendix C, Figure C4-32).

Level of Evidence in the Health Effect. Using the OHAT method (see Figure 3-3), a very low confidence rating in the body of evidence and questionable evidence of an effect result in a conclusion that there is an inadequate level of evidence to assess whether fetal exposure to DINP is associated with a decrease in AGD in male rats.

A summary of the confidence ratings of all the other phthalates and effects on AGD is presented in an evidence profile in Table 3-15.

TABLE 3-14 Studies of DINP and AGD in Rats

NOTES: The earliest life stage evaluated is shown when multiple observation times were assessed. GD, gestation day; LOAEL, lowest-observed-adverse-effect level; NOAEL, no-observed-adverse-effect level; PND, postnatal day.

TABLE 3-15 Profile of the Confidence in the Body of Evidence on BzBP, DBP, and DINP and AGD in Animals

NOTE: Evidence on DEHP included for comparison.

^a Ashby et al. (1997); Nagao et al. (2000); Ema and Miyawaki (2002); Tyl et al. (2004); Aso et al. (2005); Ahmad et al. (2014).

^b Ema et al. (1998, 2000); Mychreest et al. (1998, 1999, 2000); Wolfe and Patel (2002); Barlow et al. (2004); Lee et al. (2004); Zhang et al. (2004); Jiang et al. (2007): Drake et al. (2009); Li et al. (2009); Martino-Andrade et al. (2009); Struve et al. (2009); Kim et al. (2010); MacLeod et al. (2010); Scarano et al. (2010); Johnson et al. (2011); van den Driesche et al. (2012); Ahmad et al. (2014); Giribabu et al. (2014); N. Li et al. (2015).

^c Masutomi et al. (2003); Boberg et al. (2011); Clewell et al. (2013); L. Li et al. (2015).

^d Moore et al. (2001); Borch et al. (2004); Jarfelt et al. (2005); Wolfe and Layton (2005); Andrade et al. (2006); Culty et al. (2008); Lin et al. (2008, 2009); Christiansen et al. (2009, 2010); Gray et al. (2009); Martino-Andrade et al. (2009); Vo et al. (2009); Li et al. (2013); Zhang et al. (2013); Jones et al. (2015).

^e Liu et al. (2008); Do et al. (2012); Pocar et al. (2012).

Fetal Testosterone Concentrations

BzBP and Fetal Testosterone Concentrations

Confidence in the Body of Evidence. There is high confidence in the body of evidence on developmental exposure to BzBP and effects on fetal testosterone in rats. Two studies examining BzBP and fetal testosterone in rats were available (see Table 3-16). No significant risk of bias concerns were found (both studies controlled for litter effects) or other factors that would warrant a downgrade. One study found a greater than 50% decrease in fetal testosterone in rats given BzBP at ≥100 mg/kg-day on GD 14-18 (Furr et al. 2014). Another study found a 22-90% decrease in fetal testosterone in rats given BzBP at 300-900 mg/kg-day on GD 8-18 (Howdeshell et al. 2008). Confidence in the BzBP evidence was therefore upgraded on two factors because both studies showed a relatively large magnitude of change and reflected a similar magnitude of response within the same dose range (see Appendix C, Figures C4-15 and C4-16).

Level of Evidence in the Health Effect. Using the OHAT method (see Figure 3-3), a high confidence rating in the body of evidence and evidence of an effect result in a conclusion that there is a high level of evidence that fetal exposure to BzBP is associated with a decrease in fetal testosterone in male rats.

DBP and Fetal Testosterone Concentrations

Confidence in the Body of Evidence. There is high confidence in the body of evidence on developmental exposure to DBP and effects on fetal testosterone in rats. Twelve studies examining DBP and fetal testosterone in rats were available (see Table 3-17). Overall, the risk of bias concerns did not warrant a downgrade (see Figure 3-22). Multiple studies found a greater than 40% decrease in fetal testosterone in rats given DBP at ≥100 mg/kg-day during gestation. Confidence in the DBP evidence was upgraded on two factors because the studies showed a relatively large magnitude of change and reflected a similar magnitude of response within the same dose range (Appendix C, Figures C4-23 and C4-24).

Level of Evidence in the Health Effect. Using the OHAT method (see Figure 3-3), a high confidence rating in the body of evidence and evidence of an effect result in a conclusion that there is a high level of evidence that fetal exposure to DBP is associated with a decrease in fetal testosterone in male rats.

TABLE 3-16 Studies of BzBP and Fetal Testosterone in Rats

* Indicates that testes testosterone production was measured.

NOTES: The earliest life stage evaluated is shown when multiple observation times were assessed. GD, gestation day; LOAEL, lowest-observed-adverse-effect level; NOAEL, no-observed-adverse-effect level; PND, postnatal day.

TABLE 3-17 Studies of DBP and Fetal Testosterone in Rats

* Indicates that testes testosterone production was measured.

NOTES: The earliest life stage evaluated is shown when multiple observation times were assessed. GD, gestation day; LOAEL, lowest-observed-adverse-effect level; NOAEL, no-observed-adverse-effect level; PND, postnatal day.

DIBP and Fetal Testosterone Concentrations

Confidence in the Body of Evidence. There is high confidence in the body of evidence on developmental exposure to DIBP and effects on fetal testosterone in rats. Two studies examining DBP and fetal testosterone in rats were available (see Table 3-18). Confidence in the DIBP evidence was not downgraded because of risk of bias concerns (see Figure 3-23). Confidence was downgraded because of imprecision detected in a meta-analysis of the studies (see discussion of the meta-analysis later in the chapter). Both studies found a decrease in fetal testosterone of 40% or more in rats given DIBP at ≥100 mg/kg-day during gestation (Howdeshell et al. 2008; Hannas et al. 2011b). Confidence in the DIBP evidence was upgraded for two factors because these studies showed a relatively large magnitude of change and reflected a similar magnitude of response within the same dose range (see Appendix C, Figures C4-29 and C4-30).

Level of Evidence in the Health Effect. Using the OHAT method (see Figure 3-3), a high confidence rating in the body of evidence and evidence of an effect result in a conclusion that there is a high level of evidence that fetal exposure to DIBP is associated with a decrease in fetal testosterone in male rats.

TABLE 3-18 Studies of DIBP and Fetal Testosterone in Rats

* Indicates that testes testosterone production was measured.

NOTES: The earliest life stage evaluated is shown when multiple observation times were assessed. GD, gestation day; LOAEL, lowest-observed-adverse-effect level; NOAEL, no-observed-adverse-effect level; PND, postnatal day.

DINP and Fetal Testosterone Concentrations

Confidence in the Body of Evidence. There is high confidence in the body of evidence on developmental exposure to DINP and effects on fetal testosterone in rats. Four studies examining DINP and fetal testosterone in rats were available (see Table 3-19). Confidence in the evidence was not downgraded because of risk of bias concerns (see Figure 3-24). Confidence was downgraded because of imprecision detected in a meta-analysis of the studies (see discussion of the meta-analysis later in the chapter). Confidence in the DINP evidence was upgraded on two factors because these studies reflected a large magnitude of effect and a similar magnitude of response within the same dose range (see Appendix C, Figures C4-29 and C4-30).

TABLE 3-19 Studies of DINP and Fetal Testosterone in Rats

* Indicates that testes testosterone production was measured.

NOTES: The earliest life stage evaluated is shown when multiple observation times were assessed. GD, gestation day; LOAEL, lowest-observed-adverse-effect level; NOAEL, no-observed-adverse-effect level; PND, postnatal day.

Level of Evidence in the Health Effect. Using the OHAT method (see Figure 3-3), a high confidence rating in the body of evidence and evidence of an effect result in a conclusion that there is a high level of evidence that fetal exposure to DINP is associated with a decrease in fetal testosterone in male rats.

DPP and Fetal Testosterone Concentrations

Confidence in the Body of Evidence. There is high confidence in the body of evidence on developmental exposure to DPP and effects on fetal testosterone in rats. Four studies examining DPP and fetal testosterone in rats were available (see Table 3-20). Confidence in the evidence was not downgraded because of risk of bias concerns (see Figure 3-25). Confidence in the evidence was upgraded on two factors because these studies showed a relatively large magnitude of change and reflected a similar magnitude of response within the same dose range (see Appendix C, Figures C4-38 and C4-39).

TABLE 3-20 Studies of DPP and Fetal Testosterone in Rats

* Indicates that testes testosterone production was measured.

** Multiple blocks in this experiment.

NOTES: The earliest life stage evaluated is shown when multiple observation times were assessed. GD, gestation day; LOAEL, lowest-observed-adverse-effect level; NOAEL, no-observed-adverse-effect level; PND, postnatal day.

Level of Evidence in the Health Effect. Using the OHAT method (see Figure 3-3), a high confidence rating in the body of evidence and evidence of an effect result in a conclusion that there is a high level of evidence that fetal exposure to DPP is associated with a decrease in fetal testosterone in male rats.

A summary of the confidence ratings on all the phthalates and effects on fetal testosterone in animals is presented in an evidence profile in Table 3-21.

BzBP and Hypospadias

Confidence in the Body of Evidence. There is moderate confidence in the body of evidence on BzBP and hypospadias. Two studies examining BzBP and hypospadias in rats were found (see Table 3-22), one of which reported an increased (but not statistically significant) incidence. Confidence in the evidence was downgraded because of risk of bias concerns (see Figure 3-26); both studies had a probably high risk of bias rating because of concerns about whether the researchers were blinded to the treatment groups and concerns about the outcome measures, and one study did not control for litter effects. Because the data are limited and there were risk of bias concerns regarding the outcome measure, the committee did not upgrade confidence in the body of evidence for the finding of a rare effect as discussed earlier with DEHP phthalates and hypospadias.

Level of Evidence in the Health Effect. Using the OHAT method (see Figure 3-3), a moderate confidence rating in the body of evidence and evidence of an effect result in a conclusion that there is a moderate level of evidence that fetal exposure to BzBP is associated with an increase in hypospadias in rats.

TABLE 3-21 Profile of the Confidence in the Body of Evidence on BzBP, DBP, DIBP, DINP, and DPP and Fetal Testosterone Concentrations in Animals

NOTE: Evidence on DEHP included for comparison.

^a Howdeshell et al. (2008); Furr et al. (2014).

^b Lehmann et al. (2004); Johnson et al. (2007, 2011); Kuhl et al. (2007); Mahood et al. (2007); Howdeshell et al. (2008); Clewell et al. (2009); Matrino-Andrade et al. (2009); Struve et al. (2009); van den Driesche et al. (2012); Furr et al. (2014); N. Li et al. (2015).

^c Howdeshell et al. (2008); Hannas et al. (2011b).

^d Adamson et al. (2009); Boberg et al. (2011); Hannas et al. (2011b); L. Li et al. (2015).

^e Howdeshell et al. (2008); Hannas et al. (2011a); Beverly et al. (2014); Furr et al. (2014).

^f Do et al. (2012).

^g Borch et al. (2004, 2006); Culty et al. (2008); Howdeshell et al. (2008); Lin et al. (2008); Martino-Andrade et al. (2009); Vo et al. (2009); Hannas et al. (2011b); Klinefelter et al. (2012); Saillenfait et al. (2013a); Furr et al. (2014).

TABLE 3-22 Studies of BzBP and Hypospadias in Rats

NOTE: GD, gestation day; LOAEL, lowest-observed-adverse-effect level; NOAEL, no-observed-adverse-effect level; PND, postnatal day.

DBP and Hypospadias

Confidence in the Body of Evidence. There is high confidence in the body of evidence on developmental exposure to DBP and hypospadias in rats. Eight studies examining DBP and hypospadias in rats were available (see Table 3-23). Confidence in the DBP evidence was downgraded because of risk of bias concerns (see Figure 3-27) that related to blinding of investigators and confidence in the outcome assessment. Confidence in the DPB evidence was upgraded because of a large magnitude of effect (see Appendix C, Figure C4-26) and a similar magnitude of response within the same dose range (see Appendix C, Figure C4-27). Confidence in the body of evidence was also upgraded because the background control incidence of hypospadias was reported as zero across all studies, so any positive finding was considered treatment related (i.e., rare outcome).

Level of Evidence in the Health Effect. Using the OHAT method (see Figure 3-3), a high confidence rating in the body of evidence and evidence of an effect result in a conclusion that there is a high level of evidence that fetal exposure to DBP is associated with an increase in hypospadias in rats.

A summary of the confidence ratings of all the phthalates and hypospadias in rats is presented in an evidence profile in Table 3-24.

Meta-Analyses of Animal Data

The animal database for AGD and BzBP and DBP were judged to be amenable calculated meta-analysis. Similar methods were used as previously described for DEHP. BMD₅ estimates were using a linear or linear-quadratic model, with the model selection based on the lowest AICc. The BMD₅ was calculated only for the “fixed effect”—that is, the estimated mean response across studies.

For AGD, there were statistically significant overall effects and linear trends in log₁₀(dose) and dose for both BzBP and DBP. The statistical significance of these effects was robust to leaving out individual studies and restricting to the highest dose group from each study. A summary of the analysis is provided in Table 3-25, and supporting details are presented in Appendix C, Section C-6.

TABLE 3-23 Studies of DBP and Hypospadias in Rats

* Indicates that testes testosterone production was measured.

NOTES: The earliest life stage evaluated is shown when multiple observation times were assessed. GD, gestation day; LOAEL, lowest-observed-adverse-effect level; NOAEL, no-observed-adverse-effect level; PND, postnatal day.

TABLE 3-24 Profile of the Confidence in the Body of Evidence on BzBP and DBP and Hypospadias in Animals

NOTE: Evidence on DEHP included for comparison.

^a Nagao et al. (2000); Tyl et al. (2004).

^b Mychreest et al. (1998, 1999, 2000); Barlow et al. (2004); Jiang et al. (2007); Drake et al. (2009); Kim et al. (2010); N. Li et al. (2015).

^c Liu et al. (2008).

^d Jarfelt et al. (2005); Andrade et al. (2006); Christiansen et al. (2009, 2010); Gray et al. (2009); Saillenfait et al. (2009); Vo et al. (2009); Li et al. (2013).

TABLE 3-25 Summary of Meta-Analyses for BzBP and DBP Effects on Rat AGD

NOTE: See Appendix C-6 for additional details. Results of the meta-analyses for DEHP (all rat strains) are included for comparison.

The animal database for fetal testosterone and BzBP, DBP, DIBP, DINP, and DPP were also judged to be amenable for meta-analysis. Similar methods were used as previously described for DEHP. BMD₅ and BMD₄₀ estimates were calculated using a linear or linear-quadratic model, with the model selection based on the lowest AICc. The BMDs were calculated only for the “fixed effect”—that is, the estimated mean response across studies.

For fetal testosterone, there were statistically significant overall effects and linear trends in log10(dose) and dose for BzBP, DBP, DIBP, DINP, and DPP. The statistical significance of these effects was generally robust to leaving out individual studies and restricting to the highest dose group from each study. In the case of DIBP, there were too few studies to conduct this sensitivity analyses. A summary of the analysis is provided below (see Table 3-26), and supporting details are presented in Appendix C, Section C-6.

There were insufficient studies to perform a meta-analysis on other phthalates and hypospadias.

Human Health Effect Results on Other Phthalates

As with DEHP, the relevant human studies used state-of-the-art analytical chemistry methods to measure urinary phthalate metabolites and included collection of and adjustment for important potential confounding variables, such as measures of urinary dilution, infant body size, and maternal demographic factors. The key risk of bias evaluation factors for the human studies were whether the study designs or analyses accounted for important confounding and modifying variables, exposure characterization, and outcome assessment. There was no evidence of publication bias (see Appendix D, Table D3-1). Most studies that measured multiple urinary phthalate metabolites also measured DEHP metabolites and have been discussed previously. The initial rating for the confidence in the human studies was moderate based on the following three criteria: exposures occurred prior to outcome, outcomes were measured on individuals, and a (control) comparison group was used (see Figure 3-2 for OHAT method for rating confidence). When appropriate, meta-analyses of the human data were performed and provided additional information concerning the confidence ratings.

Reductions in AGD

Confidence in the Body of Evidence. There is moderate confidence in the body of evidence on the other phthalates and effects on AGD. Some of the prospective cohort studies of pregnant mothers and their infants that were evaluated for DEHP effects also included evaluation of association of AGD (ap or as) outcomes and other phthalate biomarkers (see Table 3-27). The relevant studies used state-of-the-art analytical chemistry methods and adjustment for important potential confounding variables. As was the case with DEHP, the effect estimates of AGD are estimates of slopes that assume a monotonic dose-response relationship between exposure and effect. Table 3-27 presents the level of confidence in the evidence for phthalate metabolites (MBP, MBzP, MCNP, MCOP, MCPP, MEP, MIBP, and MMP) and AGD in humans. The committee found no reason to upgrade or downgrade confidence in the evidence (see Appendix D, Section D-3 for details). No significant risk of bias concerns were found (see Figure 3-28).

TABLE 3-26 Summary of the Meta-Analyses for BzBP, DBP, DIBP, DINP, and DPP Effects on Rat Fetal Testosterone

* The 5% change was well below the range of the data, but it will be 10 times lower because a linear model was used.

NOTE: See Appendix C-6 for additional details. Results of the meta-analyses for DEHP (all rat strains) are included for comparison.

TABLE 3-27 Profile of the Confidence in the Body of Evidence on Phthalates and AGD in Humans

^a Swan (2008); Bornehag et al. (2015); Swan et al. (2015); and Jensen et al. (2016).

^b Swan et al. (2008); Bustamante-Montes et al. (2013); Bornehag et al. (2015); Swan et al. (2015); Jensen et al. (2016); and Martino-Andrade et al. (2016).

^c Swan (2008); Swan et al. (2015); and Jensen et al. (2016).

^d Swan et al. (2015).

^e Bornehag et al. (2015); Swan et al. (2015); and Jensen et al. (2016).

^f Swan (2008).

^g Swan (2008); Swan et al. (2015).

Meta-Analyses of Human Data on AGD and BzBP, DBP, DEP, DIBP, and DINP. Meta-analyses of human studies on BzBP, DBP, DEP, DIBP, and DINP in relation to alterations in AGD were conducted (see Appendix D, Section D-5). The same meta-analysis methods used for DEHP were applied to these phthalates. Three phthalates—DIDP, DMP, and DOP—had only one study precluding conduct of meta-analyses for these phthalates. As with the DEHP analysis, AGD (as) is preferred over AGD (ap) for each study. For the studies by Bustamonte-Montes et al. (2013) and Swan (2008), the confidence interval was estimated using the reported p-value, assuming a normal distribution. Sensitivity analyses included leaving one study out at a time and using AGD (ap) exclusively as the outcome measure. Beta coefficients standardized to a percent change per log₁₀ change in metabolite exposure were used. A summary of these meta-analyses and an interpretation of the results are provided in Table 3-28.

TABLE 3-28 Summary of Meta-Analyses of Human Studies of BzBP, DBP, DEP, DIBP, DINP and AGD

NOTE: Overall pooled estimate from random effects (RE) model per 10-fold increase in metabolite exposure is provided.

Level of Evidence in the Health Effect. Meta-analyses of the AGD studies on DBP and DEP found some evidence of a decrease in AGD being associated with exposure to these phthalates. The results show a consistent pattern of findings that higher maternal urinary concentrations of DBP and DEP during pregnancy are associated with a reduction in AGD in infancy. The small amount of heterogeneity observed for DEP may be due to sample size differences, AGD measurement variability, urinary metabolite concentration variability, and the potential for residual confounding. Using the OHAT method (see Figure 3-3), a moderate confidence rating in the body of evidence and evidence of an effect result in a conclusion that there is a moderate level of evidence that fetal exposure to DBP and DEP are associated with a reduction in AGD in male infants.

Meta-analyses of the AGD studies on BzBP, DIBP, and DINP do not show evidence of an effect. Thus, a moderate confidence rating in the body of evidence and no evidence of an effect results in a conclusion that there is an inadequate level of evidence to assess whether fetal exposure to these phthalates is associated with a decrease in AGD in male infants.

Fetal Testosterone Concentrations

Given the disparate matrices used to measure testosterone (amniotic fluid, maternal serum, or cord blood), the differences in timing of exposure (during pregnancy or at delivery), and the paucity of studies, the committee decided the data were insufficient to pursue further analysis of effects of other phthalate metabolites on fetal testosterone. (See the earlier discussion on DEHP for further details.) The committee concluded that there is inadequate evidence to determine whether fetal exposure to BzBP, DBP, DEP, DIBP, DIDP, DINP, DMP, and DOP is associated with a reduction in fetal testosterone in male humans.

Hypospadias

Given the disparate matrices used to measure phthalate metabolites (amniotic fluid and urine) and the paucity of studies, the committee decided the data were insufficient to pursue further analysis of effects of other phthalate metabolites on hypospadias. (See the earlier discussion on DEHP for further details.) The committee concluded that there is inadequate evidence to determine whether fetal exposure to BzBP, DBP, DEP, DIBP, DIDP, DINP, DMP, and DOP is associated with a hypospadias in humans.

Summary of Initial Hazard Evaluations for Other Phthalates

Table 3-29 provides initial hazard evaluations for other phthalates and AGD in humans based on the OHAT hazard identification scheme (see Figure 3-4).

RELEVANCE TO ANIMAL TOXICITY TESTING

The committee’s systematic reviews and evidence integration illustrate the use of its iterative strategy for evaluating low-dose responses in humans. The rodent and human studies reviewed by the committee shared common outcome measures, AGD, changes in testosterone, and hypospadias. Despite this apparent similarity in outcomes, significant differences are present in the two data streams. The animal studies reviewed by the committee examined single phthalate exposures, whereas the human epidemiologic studies involved subjects exposed to multiple phthalates. These differences highlight the need for additional focused animal studies that more closely mimic human exposures. In addition, the human epidemiologic studies relied on biomarker data (e.g., analysis of DEHP metabolite concentrations in maternal urine during gestation as an estimate of internal dose), whereas the animal studies relied on DEHP external dose. The application of animal data to risk assessment would be strengthened by the inclusion of pharmacokinetic evaluations that could yield comparable biomonitoring data for animals that are collected in people. Despite these limitations, this case represents an example where current toxicity-testing paradigms can detect a hazard that is presumed to be of concern to humans but might not be accurately

predicting doses at which effects occur in humans. It also provides additional support for prior EPA decisions to include AGD measurements in regulatory toxicology testing (Chapter 2, Box 2-5).

Table 3-30 provides initial hazard evaluations for other phthalates and fetal testosterone in humans based on the OHAT hazard identification scheme (see Figure 3-4). Table 3-31 provides initial hazard evaluations for other phthalates and hypospadias in humans based on the OHAT hazard identification scheme (see Figure 3-4).

TABLE 3-29 Initial Hazard Evaluations for Other Phthalates and AGD in Humans

TABLE 3-30 Initial Hazard Evaluations for Other Phthalates and Fetal Testosterone in Humans

TABLE 3-31 Initial Hazard Evaluations for Other Phthalates and Hypospadias in Humans

FINDINGS AND RECOMMENDATIONS

Systematic Reviews

• Consistency and Transparency: The committee found that the systematic review process was valuable because it provided a framework for identifying, selecting, and evaluating evidence in a consistent and explicit manner; maximized transparency in how the assessments were performed; and facilitated the clear presentation of the basis for scientific judgments.
• Meta-analyses: The committee found that the meta-analyses were valuable in summarizing data from the systematic reviews and in comparing the animal and the human evidence in a robust and consistent manner. For example, the meta-analyses of the animal and the human studies on

• DEHP (and select other phthalates) and AGD, and of the animal studies on DEHP and (select other phthalates) and fetal testosterone, provided quantitative evidence that certain phthalates are associated with reductions in AGD and in fetal testosterone concentrations. The meta-analysis results not only informed the confidence ratings of the body of evidence but also allowed the committee to estimate benchmark doses on the basis of data from multiple studies.

  Recommendation: Systematic reviews should include meta-analysis of the animal and the human evidence, if appropriate. The results of meta-analyses should be used to examine quantitative relationships between EACs and end points of interest, to inform the confidence ratings of the bodies of evidence, and, if possible, to estimate benchmark doses.

• Risk of Bias Evaluations: Information important to the evaluation of the quality of individual animal studies was often not reported, including whether the study controlled for litter effects, whether animals were randomly allocated to study groups, and whether research personnel were blinded to the study groups during the outcome assessment. Because a lack of adequate reporting could not be distinguished from failure to adhere to practices that minimize bias, failure to report practices that minimize bias often led to higher risk of bias ratings for individual studies, downgrading the overall level of confidence in the body of evidence. These types of problems could be remedied if journals required better reporting of the methods used in animal studies, especially reporting pertaining to issues that might introduce bias into the research. These requirements could build on reporting standards that have been developed by various organizations to improve transparency (e.g., the ARRIVE guidelines [Kilkenny et al. 2010]). For example, studies should be required to report whether animals were assigned to study groups using random allocation and whether researchers were blinded to the study groups during outcome assessment.

Evidence Integration

• A comparison of doses between animal and human studies was challenging and imprecise because animal studies characterized exposure as the administered dose (the amount of chemical that was fed or otherwise administered), whereas the human studies measured internal dose (a measurement of the chemical in a biological sample). There is some indication that the difference in internal dose between humans and rodents may be less than the difference in administered dose; these estimates are uncertain, however, and additional work is needed for clarification. Toxicology studies that measure internal dose metrics, especially with the same measure used in human biomonitoring, could help address this data gap.

  Recommendation: To support animal-to-human extrapolations, pharmacokinetic data should be generated and used to develop pharmacokinetic models that make it possible to infer human internal doses (not just intake) from biomonitoring data and animal internal doses from administered doses.

• For the evaluation of the effect of phthalates on testosterone concentrations, it was difficult to integrate findings because human studies relied on surrogate measures of in utero concentrations (testosterone measured in maternal urine, amniotic fluid, or cord serum), whereas most animal studies measured fetal testosterone in the testes of rodents. Targeted animal studies that evaluate the relationship between phthalate exposure and changes in testosterone concentrations in biological matrices more relevant to measures taken in human studies could help address this data gap.

Mechanistic Information

• Mechanistic data from animal and xenograft studies were available, and adverse outcome pathways have been proposed for how exposure to phthalates during the male programming window affects reproductive-tract development. The committee found that this information was useful during the scoping and problem formulation phase of the systematic review to help determine what outcomes should be the focus and how evidence could be integrated in reaching conclusions.
• The results of the meta-analyses and subsequent benchmark dose analyses of experimental animal data on two end points hypothesized to be part of an adverse outcome pathway (AGD and fetal testosterone concentrations) revealed species- and strain-specific quantitative differences that were not entirely consistent. For instance, there was some evidence of reduced AGD in mice exposed to DEHP, but without evidence of decreased testosterone (or increased hypospadias). Additionally, compared to Sprague-Dawley rats, Wistar rats appeared to be more sensitive to changes in AGD but less sensitive to decreases in testosterone and much less sensitive to hypospadias.
• Given the variation in sensitivities among rodent species and strains, the mechanistic information was less useful in considering concordance and discordance between the animal and the human studies beyond providing evidence of the biological plausibility that the effects observed in rat studies identify the same hazards in humans.
• A meta-analysis of xenograft data on DBP and serum testosterone was performed to show how meta-analyses could be applied to mechanistic studies.

Hazard Identification

• The committee concluded that exposure to DEHP and certain other phthalates are presumed to cause decreases in AGD in humans, based on a moderate level of evidence from human studies and a high level of evidence from animal studies. Measurement of AGD in rodents has value in the identification of reproductive hazards in humans associated with EACs.
• The committee concluded that exposure to DEHP and certain other phthalates are presumed to cause decreases in fetal testes testosterone, based on inadequate evidence from human studies and a high level of evidence from animal studies.
• The committee concluded that exposure to DEHP and certain other phthalates is suspected to cause increases in the risk of hypospadias in humans, based on inadequate evidence from human studies and a moderate level of evidence from animal studies.

Low-Dose Effects

• The committee concluded that the human studies provide a moderate level of evidence that fetal exposure to DEHP is associated with decreases in AGD in humans. Uncertainty in the internal doses of humans relative to experimental animals limited the ability to draw conclusions about the prediction of low-dose effects based on experimental animal studies. The development of pharmacokinetic data and models for extrapolation of data from animal studies or human biomonitoring data could facilitate the evaluation of the potential of phthalates to cause health effects in humans at low doses.

Other Considerations

• Mixtures: The committee found that humans are exposed to a mixture of phthalates, whereas the experimental animal evidence was generally from studies with single phthalate congeners.

• This difference between human mixture exposure and the single chemical animal exposures contributed to the challenges for integrating evidence between human and animal studies.

• Expertise: The committee found that the conduct of the systematic review and evidence integration requires a multidisciplinary approach that should be tailored to the specific review question. Experts in the conduct of meta-analyses and benchmark dose modeling will be essential.

REFERENCES

Borch, J., S.B. Metzdorff, A.M. Vinggaard, L. Brokken, and M. Dalgaard. 2006. Mechanisms underlying the antiandrogenic effects of diethylhexyl phthalate in fetal rat testis. Toxicology 223(1-2):144-155.

Bornehag, C.G., F. Carlstedt, B.A. Jonsson, C.H. Lindh, T.K. Jensen, A. Bodin, C. Jonsson, S. Janson, and S.H. Swan. 2015. Prenatal phthalate exposures and anogenital distance in Swedish boys. Environ. Health Perspect. 123(1):101-107.

Bustamante-Montes, L.P., M.A. Hernández-Valero, D. Flores-Pimentel, M. García-Fábila, A. Amaya-Chávez, D.B. Barr, and V.H. Borja-Aburto. 2013. Prenatal exposure to phthalates is associated with decreased anogenital distance and penile size in male newborns. J. Dev. Orig. Health Dis. 4(4):300-306.

Calafat, A.M., J.W. Brock, M.J. Silva, L.E. Gray, Jr., J.A. Reidy, D.B. Barr, and L.L. Needham. 2006. Urinary and amniotic fluid levels of phthalate monoesters in rats after the oral administration of di(2-ethylhexyl) phthalate and di-n-butyl phthalate. Toxicology 217(1):22-30.

Calafat, A.M., M.P. Longnecker, H.M. Koch, S.H. Swan, R. Hauser, L.R. Goldman, B.P. Lanphear, R.A. Rudel, S.M. Engel, S.L. Teitelbaum, R.M. Whyatt, and M.S. Wolff. 2015. Optimal exposure biomarkers for nonpersistent chemicals in environmental epidemiology. Environ. Health Perspect. 123(7):A166-A168.

CDC (Centers for Disease Control and Prevention). 2009. Fourth Report on Human Exposure to Environmental Chemicals. Atlanta, GA: U.S. Department of Health and Human Services, Centers for Disease Control and Prevention [online]. Available: http://www.cdc.gov/exposurereport/ [accessed June 15, 2016].

CDC. 2015. Fourth Report on Human Exposure to Environmental Chemicals, Updated Tables, February 2015. Atlanta, GA: U.S. Department of Health and Human Services, Centers for Disease Control and Prevention [online]. Available: http://www.cdc.gov/exposurereport/ [access June 15, 2016].

Chang, J.W., C.C. Lee, W.H. Pan, W.C. Chou, H.B. Huang, H.C. Chiang, and P.C. Huang. 2017. Estimated daily intake and cumulative risk assessment of phthalates in the general Taiwanese after the 2011 DEHP food scandal. Sci. Rep. 7:45009.

CHAP (Chronic Hazard Advisory Panel). 2014. Report to the U.S. Consumer Product Safety Commission by the Chronic Hazard Advisory Panel on Phthalates and Phthalate Alternatives. U.S. Consumer Product Safety Commission, Directorate for Health Sciences, Bethesda, MD [online]. Available: https://www.cpsc.gov/PageFiles/169902/CHAP-REPORT-With-Appendices.pdf [accessed June 17, 2016].

Chauvigné, F., A. Menuet, L. Lesné, M.C. Chagnon, C. Chevrier, J.F. Regnier, J. Angerer, and B. Jégou. 2009. Time- and dose-related effects of di-(2-ethylhexyl) phthalate and its main metabolites on the function of the rat fetal testis in vitro. Environ. Health Perspect. 117(4):515-521.

Chevrier, C., C. Petit, C. Philippat, M. Mortamais, R. Slama, F. Rouget, A.M. Calafat, X. Ye, M.J. Silva, M.A. Charles, and S. Cordier. 2012. Maternal urinary phthalates and phenols and male genital anomalies. Epidemiology 23(2):353-356.

Christiansen, S., M. Scholze, M. Dalgaard, A.M. Vinggaard, M. Axelstad, A. Kortenkamp, and U. Hass. 2009. Synergistic disruption of external male sex organ development by a mixture of four antiandrogens. Environ. Health Perspect. 117(12):1839-1846.

Christiansen, S., J. Boberg, M. Axelstad, M. Dalgaard, A.M. Vinggaard, S.B. Metzdorff, and U. Hass. 2010. Low-dose perinatal exposure to di(2-ethylhexyl) phthalate induces anti-androgenic effects in male rats. Reprod. Toxicol. 30(2):313-321.

Clewell, R.A., J.J. Kremer, C.C. Williams, J.L. Campbell, M.A. Sochaski, M.E. Andersen, and S.J. Borghoff. 2009. Kinetics of selected di-n-butyl phthalate metabolites and fetal testosterone following repeated and single administration in pregnant rats. Toxicology 255(1-2):80-90.

Clewell, R.A., J.L. Campbell, S.M. Ross, K.W. Gaido, H.J. Clewell, III, and M.E. Andersen. 2010. Assessing the relevance of in vitro measures of phthalate inhibition of steroidogenesis for in vivo response. Toxicol. In Vitro 24(1):327-334.

Clewell, R.A., A. Thomas, G. Willson, D.M. Creasy, and M.E. Andersen. 2013. A dose response study to assess effects after dietary administration of diisononyl phthalate (DINP) in gestation and lactation on male rat sexual development. Reprod. Toxicol. 35:70-80.

Culty, M., R. Thuillier, W. Li, Y. Wang, D.B. Martinez-Arguelles, C.G. Benjamin, K.M. Triantafilou, B.R. Zirkin, and V. Papadopoulos. 2008. In utero exposure to di-(2-ethylhexyl) phthalate exerts both short-term and long-lasting suppressive effects on testosterone production in the rat. Biol. Reprod. 78(6):1018-1028.

Dean, A., and R.M. Sharpe. 2013. Clinical review: Anogenital distance or digit length ratio as measures of fetal androgen exposure: Relationship to male reproductive development and its disorders. J. Clin. Endocrinol. Metab. 98(6):2230-2238.

Do, R.P., R.W. Stahlhut, D. Ponzi, F.S. vom Saal, and J.A. Taylor. 2012. Non-monotonic dose effects of in utero exposure to di(2-ethylhexyl) phthalate (DEHP) on testicular and serum testosterone and anogenital distance in male mouse fetuses. Reprod. Toxicol. 34(4):614-621.

Drake, A.J., S. van den Driesche, H.M. Scott, G.R. Hutchison, J.R. Seckl, and R.M. Sharpe. 2009. Glucocorticoids amplify dibutyl phthalate-induced disruption of testosterone production and male reproductive development. Endocrinology 150(11):5055-5064.

Ema, M., and E. Miyawaki. 2002. Effects on development of the reproductive system in male offspring of rats given butyl benzyl phthalate during late pregnancy. Reprod. Toxicol. 16(1):71-76.

Ema, M., E. Miyawaki, and K. Kawashima. 1998. Further evaluation of developmental toxicity of di-n-butyl phthalate following administration during late pregnancy in rats. Toxicol. Lett. 98(1-2):87-93.

Ema, M., E. Miyawaki, and K. Kawashima. 2000. Critical period for adverse effects on development of reproductive system in male offspring of rats given di-n-butyl phthalate during late pregnancy. Toxicol. Lett. 111(3):271-278.

EU (European Union). 2004. Commission Directive 2004/93/EC of 21 September 2004 amending Council Directive 76/768/EEC for the purpose of adapting its Annexes II and III to technical progress. O.J. EU I.300:13-41. September 25, 2004 [online]. Available: http://eur-lex.europa.eu/LexUriServ/LexUriServ.do?uri=OJ:L:2004:300:0013:0041:EN:PDF [accessed February 13, 2017].

EU. 2005a. Directive 2005/84/EC of the European Parliament and of the Council 14 December 2005 amending for the 22nd time Council Directive 76/769/EEC on the approximation of the laws, regulations and administrative provisions of the member states relating to restrictions on the marketing and use of certain dangerous substances and preparations (phthalates in toys and childcare articles). O.J. EU L344:40-43. December 27, 2005 [online]. Available: http://eur-lex.europa.eu/LexUriServ/LexUriServ.do?uri=OJ:L:2005:344:0040:0043:EN:PDF [accessed February 13, 2107].

EU. 2005b. Commission Directive 2005/80/EC of 21 November 2005 amending Council Directive 76/768/EEC, concerning cosmetic products, for the purposes of adapting Annexes II and III there to technical progress. O.J. EU I.303:32-37. November 22, 2005 [online]. Available: http://eur-lex.europa.eu/legal-content/EN/TXT/PDF/?uri=CELEX:32005L0080&from=EN [accessed February 13, 2017].

EU. 2007. Commission Directive 2007/19/EC of 30 March 2007 amending Directive 2002/72/EC relating to plastic materials and articles intended to come into contact with food and Council Directive 85/572/EEC laying down the list of simulants to be used for testing migration of constituents of plastic materials and articles intended to come into contact with foodstuffs. O. J. EU L91:17-36. March 31, 2007 [online]. Available: http://eur-lex.europa.eu/LexUriServ/LexUriServ.do?uri=OJ:L:2007:091:0017:0036:EN:PDF [accessed February 13, 2017].

Fennell, T.R., W.L. Krol, S.C. Sumner, and R.W. Snyder. 2004. Pharmacokinetics of dibutyl phthalate in pregnant rats. Toxicol. Sci. 82(2):407-418.

Fisher, J.S., S. Macpherson, N. Marchetti, and R.M. Sharpe. 2003. Human “testicular dysgenesis syndrome”: A possible model using in-utero exposure of the rat to dibutyl phthalate. Hum. Reprod. 18(7):1383-1394.

Foster, P.M., E. Mylchreest, K.W. Gaido, and M. Sar. 2001. Effects of phthalate esters on the developing reproductive tract of male rats. Hum. Reprod. Update 7(3):231-235.

Fujii, S., K. Yabe, M. Furukawa, M. Hirata, M. Kiguchi, and T. Ikka. 2005. A two-generation reproductive toxicity study of diethyl phthalate (DEP) in rats. J. Toxicol. Sci. 30(Spec.):97-116.

Furr, J.R., C.S. Lambright, V.S. Wilson, P.M. Foster, and L.E. Gray, Jr. 2014. A short-term in vivo screen using fetal testosterone production, a key event in the phthalate adverse outcome pathway, to predict disruption of sexual differentiation. Toxicol. Sci. 140(2):403-424.

Gaido, K.W., J.B. Hensley, D. Liu. D.G. Wallace. S. Borghoff, K.J. Johnson, S.J. Hall, and K. Boekelheide. 2007. Fetal mouse phthalate exposure shows that gonocyte multinucleation is not associated with decreased testicular testosterone. Toxicol. Sci. 97(2):491-503.

Giribabu, N., S.B. Sainath, and P.S. Reddy. 2014. Prenatal di-n-butyl phthalate exposure alters reproductive functions at adulthood in male rats. Environ. Toxicol. 29(5):534-544.

Gray, L.E., Jr., J. Ostby, J. Furr, M. Price, D.N. Veeramachaneni, and L. Parks. 2000. Perinatal exposure to the phthalates DEHP, BBP, and DINP, but not DEP, DMP, or DOTP, alters sexual differentiation of the male rat. Toxicol. Sci. 58(2):350-365.

Gray, L.E., Jr., N.J. Barlow, K.L. Howdeshell, J.S. Ostby, J.R. Furr, and C.L. Gray. 2009. Transgenerational effects of di(2-ethylhexyl) phthalate in the male CRL:CD(SD) rat: Added value of assessing multiple offspring per litter. Toxicol. Sci. 110(2):411-425.

Gray, L.E., Jr., J. Furr, K.R. Tatum-Gibbs, C. Lambright, H. Sampson, B.R. Hannas, V.S. Wilson, A. Hotchkiss, and P.M. Foster. 2016. Establishing the “biological relevance” of dipentyl phthalate reductions in fetal rat testosterone production and plasma and testis testosterone levels. Toxicol. Sci. 149(1):178-191.

Guyatt, G.H., A.D. Oxman, R. Kunz, J. Brozek, P. Alonso-Coello, D. Rind, P.J. Devereaux, V.M. Montori, B. Freyschuss, G. Vist, R. Jaeschke, J.W. Williams, Jr., M.H. Murad, D. Sinclair, Y. Falck-Ytter, J. Meerpohl, C. Whittington, K. Thorlund, J. Andrews, and H.J. Schunemann. 2011. GRADE guidelines 6. Rating the quality of evidence—imprecision. J. Clin. Epidemiol. 64(12):1283-1293.

Hannas, B.R., J. Furr, C.S. Lambright, V.S. Wilson, P.M. Foster, and L.E. Gray, Jr. 2011a. Dipentyl phthalate dosing during sexual differentiation disrupts fetal testis function and postnatal development of the male SpragueDawley rat with greater relative potency than other phthalates. Toxicol. Sci. 120(1):184-193.

Hannas, B.R., C.S. Lambright, J. Furr, K.L. Howdeshell, V.S. Wilson, and L.E. Gray, Jr. 2011b. Dose-response assessment of fetal testosterone production and gene expression levels in rat testes following in utero exposure to diethylhexyl phthalate, diisobutyl phthalate, diisoheptyl phthalate, and diisononyl phthalate. Toxicol. Sci. 123(1):206-216.

Hannas, B.R., C.S. Lambright, J. Furr, N. Evans, P.M. Foster, E.L. Gray, and V.S. Wilson. 2012. Genomic biomarkers of phthalate-induced male reproductive developmental toxicity: A targeted RT-PCR array approach for defining relative potency. Toxicol. Sci. 125(2):544-557.

Hauser, R., and A.M. Calafat. 2005. Phthalates and human health. Occup. Environ. Med. 62(11):806-818.

Hedges, L.V., J. Gurevitch, and P.S. Curtis. 1999. The meta-analysis of response ratios in experimental ecology. Ecology 80(4):1150-1156.

Heger, N.E., S.J. Hall, M.A. Sandrof, E.V. McDonnell, J.B. Hensley, E.N. McDowell, K.A. Martin, K.W. Gaido, K.J. Johnson, and K. Boekelheide. 2012. Human fetal testis xenografts are resistant to phthalate-induced endocrine disruption. Environ. Health Perspect. 120(80):1137-1143.

Higgins, J.P.T., and S. Green, eds. 2011. Identifying and measuring heterogeneity. Chapter 9 in Cochrane Handbook for Systematic Reviews of Interventions, Version 5.1.0 (updated March 2011). The Cochrane Collaboration [online]. Available: http://handbook.cochrane.org/chapter_9/9_5_2_identifying_and_measuring_heterogeneity.htm [accessed March 8, 2017].

Howdeshell, K.L., V.S. Wilson, J. Furr, C.R. Lambright, C.V. Rider, C.R. Blystone, A.K. Hotchkiss, and L.E. Gray, Jr. 2008. A mixture of five phthalate esters inhibits fetal testicular testosterone production in the SpragueDawley rat in a cumulative, dose-additive manner. Toxicol. Sci. 105(1):153-165.

Howdeshell, K.L., C.V. Rider, V.S. Wilson, J.R. Furr, C.R. Lambright, and L.E. Gray, Jr. 2015. Dose addition models based on biologically relevant reductions in fetal testosterone accurately predict postnatal reproductive tract alterations by a phthalate mixture in rats. Toxicol. Sci. 148(2):488-502.

Hsieh, M.H., M.L. Eisenberg, A.B. Hittelman, J.M. Wilson, G.E. Tasian, and L.S. Baskin. 2012. Caucasian male infants and boys with hypospadias exhibit reduced anogenital distance. Hum. Reprod. 27(6):1577-1580.

Huang, P.C., P.L. Kuo, Y.Y. Chou, S.J. Lin, and C.C. Lee. 2009. Association between prenatal exposure to phthalates and the health of newborns. Environ. Int. 35(1):14-20.

Huang, P.C., C.H. Tsai, W.Y. Liang, S.S. Li, W.H. Pan, H.C. Chiang. 2015. Age and gender differences in urinary levels of eleven phthalate metabolites in general Taiwanese population after a DEHP episode. PLoS ONE 10(7):e0133782.

Huang, P.C., C.H. Tsai, W.Y. Liang, S.S. Li, H.B. Huang, and P.L. Kuo. 2016. Early phthalates exposure in pregnant women is associated with alteration of thyroid hormones. PLoS ONE 11 (7):e0159398.

Ito, Y., H. Yokota, R. Wang, O. Yamanoshita, G. Ichihara, H. Wang, Y. Kurata, K. Takagi, and T. Nakajima. 2005. Species differences in the metabolism of di(2-ethylhexyl) phthalate (DEHP) in several organs of mice, rats, and marmosets. Arch. Toxicol. 79(3):147-154.

Ito, Y., M. Kamijima, C. Hasegawa, M. Tagawa, T. Kawai, M. Miyake, Y. Hayashi, H. Naito, and T. Nakajima. 2014. Species and inter-individual differences in metabolic capacity of di(2-ethylhexyl)phthalate (DEHP) between human and mouse livers. Environ. Health Prev. Med. 19(2):117-125.

Jain, V.G., and A.K. Singal. 2013. Shorter anogenital distance correlates with undescended testis: A detailed genital anthropometric analysis in human newborns. Hum. Reprod. 28(9):2343-2349.

Jarfelt, K., M. Dalgaard, U. Hass, J. Borch, H. Jacobsen, and O. Ladefoged. 2005. Antiandrogenic effects in male rats perinatally exposed to a mixture of di(2-ethylhexyl) phthalate and di(2-ethylhexyl) adipate. Reprod. Toxicol. 19(4):505-515.

Jensen, M.S., R. Anand-Ivell, B. Norgaard-Pedersen, B.A. Jonsson, J.P. Bonde, D.M. Hougaard, A. Cohen, C.H. Lindh, R. Ivell, and G. Toft. 2015. Amniotic fluid phthalate levels and male fetal gonad function. Epidemiology 26(1):91-99.

Jensen, T.K., H. Frederiksen, H.B. Kyhl, T.H. Lassen, S.H. Swan, C.G. Bornehag, N.E. Skakkebaek, K.M. Main, D.V. Lind, S. Husby, and A.M. Andersson. 2016. Prenatal exposure to phthalates and anogenital distance in male infants from a low-exposed Danish cohort (2010-2012). Environ. Health Perspect. 124(7):1107-1113.

Jiang, J., L. Ma, L. Yuan, X. Wang, and W. Zhang. 2007. Study on developmental abnormalities in hypospadiac male rats induced by maternal exposure to di-n-butyl phthalate (DBP). Toxicology 232(3):286-293.

Johns, L.E., G.S. Cooper, A. Galizia, and J.D. Meeker. 2015. Exposure assessment issues in epidemiology studies of phthalates. Environ Int. 85:27-39.

Johnson, K.J., J.B. Hensley, M.D. Kelso, D.G. Wallace, and K.W. Gaido. 2007. Mapping gene expression changes in the fetal rat testis following acute dibutyl phthalate exposure defines a complex temporal cascade of responding cell types. Biol. Reprod. 77(6):978-989.

Johnson, K.J., E.N. McDowell, M.P. Viereck, and J.Q. Xia. 2011. Species-specific dibutyl phthalate fetal testis endocrine disruption correlates with inhibition of SREBP2-dependent gene expression pathways. Toxicol. Sci. 120(2):460-474.

Johnson, K.J., N.E. Heger, and K. Boekelheide. 2012. Of mice and men (and rats): Phthalate-induced fetal testis endocrine disruption is species-dependent. Toxicol. Sci.129(2):235-248.

Jones, S., A. Boisvert, S. Francois, L. Zhang, and M. Culty. 2015. In utero exposure to di-(2-ethylhexyl) phthalate induces testicular effects in neonatal rats that are antagonized by genistein cotreatment. Biol. Reprod. 93(4):92.

Karmaus, A.L., C.M. Toole, D.L. Filer, K.C. Lewis, and M.T. Martin. 2016. High-throughput screening of chemical effects on steroidogenesis using H295R human adrenocortical carcinoma cells. Toxicol. Sci. 150(2):323-332.

Kay, V.R., M.S. Bloom, and W.G. Foster. 2014. Reproductive and developmental effects of phthalate diesters in males. Crit. Rev. Toxicol. 44(6):467-498.

Kessler, W., W. Numtip, K. Grote, G.A. Csanády, I. Chahoud, and J.G. Filser. 2004. Blood burden of di(2-ethylhexyl) phthalate and its primary metabolite mono(2-ethylhexyl) phthalate in pregnant and nonpregnant rats and marmosets. Toxicol. Appl. Pharmacol. 195(2):142-153.

Kessler, W., W. Numtip, W. Volkel, E. Seckin, G.A. Csanady, C. Putz, D. Klein, H. Fromme, and J.G. Filser. 2012. Kinetics of di(2ethylhexyl) phthalate (DEHP) and mon(2-ethylhexyl) phthalate in blood and of DEHP metabolites in urine of male volunteers after single ingestion of ring-deuterated DEHP. Toxicol. Appl. Pharmacol. 264(2):284-291.

Kilkenny, C., W.J. Browne, I.C. Cuthill, M. Emerson, and D.G. Altman. 2010. Improving bioscience research reporting: The ARRIVE guidelines for reporting animal research. PLOS Biol. 8(6):e1000412.

Kim, T.S., K.K. Jung, S.S. Kim, I.H. Kang, J.H. Baek, H.S. Nam, S.K. Hong, B.M. Lee, J.T. Hong, K.W. Oh, H.S. Kim, S.Y. Han, and T.S. Kang. 2010. Effects of in utero exposure to di(n-butyl) phthalate on development of male reproductive tracts in Sprague-Dawley rats. J. Toxicol. Environ. Health A 73(21-22):1544-1559.

Klinefelter, G.R., J.W. Laskey, W.M. Winnik, J.D. Suarez, N.L. Roberts, L.F. Strader, B.W. Riffle, and D.N. Veeramachaneni. 2012. Novel molecular targets associated with testicular dysgenesis induced by gestational exposure to diethylhexyl phthalate in the rat: A role for estradiol. Reproduction 144(6):747-761.

Koch, H.M., H.M. Bolt, and J. Angerer. 2004. Di(2-ethylhexyl)phthalate (DEHP) metabolites in human urine and serum after a single oral dose of deuterium-labelled DEHP. Arch. Toxicol. 78(3):123-130.

Kuhl, A.J., S.M. Ross, and K.W. Gaido. 2007. CCAAT/enhancer binding protein β, but not steroidogenic factor-1, modulates the phthalate-induced dysregulation of rat fetal testicular steroidogenesis. Endocrinology 148(12):5851-5864.

Kurata, Y., F. Makinodan, N. Shimamura, and M. Katoh. 2012. Metabolism of di (2-ethylhexyl) phthalate (DEHP): Comparative study in juvenile and fetal marmosets and rats. J. Toxicol. Sci. 37(1):33-49.

Lajeunesse, M.J. 2011. On the meta-analysis of response ratios for studies with correlated and multi-group designs. Ecology 92(11):2049-2055.

Lambrot, R., V. Muczynski, C. Lécureuil, G. Angenard, H. Coffigny, C. Pairault, D. Moison, R. Frydman, R. Habert, and V. Rouiller-Fabre. 2009. Phthalates impair germ cell development in the human fetal testis in vitro without change in testosterone production. Environ. Health Perspect. 117(1):32-37.

Latini, G. 2005. Monitoring phthalate exposure in humans. Clin. Chim. Acta 361(1-2):20-29.

Lee, K.Y., M. Shibutani, H. Takagi, N. Kato, S. Takigami, C. Uneyama, and M. Hirose. 2004. Diverse developmental toxicity of di-n-butyl phthalate in both sexes of rat offspring after maternal exposure during the period from late gestation through lactation. Toxicology 203(1-3):221-238.

Lehmann, K.P., S. Phillips, M. Sar, P.M. Foster, and K.W. Gaido. 2004. Dose-dependent alterations in gene expression and testosterone synthesis in the fetal testes of male rats exposed to di(n-butyl) phthalate. Toxicol. Sci. 81(1):60-68.

Lehraiki, A., C. Racine, A. Krust, R. Habert, and C. Levacher. 2009. Phthalates impair germ cell number in the mouse fetal testis by an androgen- and estrogen-independent manner. Toxicol. Sci. 111(2):372-382.

Li, H., Y. Li, J. Fang, K. Ma, S. Wu, R. Gao, X. Yan, J. Cheng, J. Liu, and F. Dong. 2014. Analysis of four phthalate monoesters in human urine using liquid chromatography–tandem mass spectrometry. LCGC North America 32(6):434-439.

Li, L., T. Bu, H. Su, Z. Chen, Y. Liang, G. Zhang, D. Zhu, Y. Shan, R. Xu, Y. Hu, J. Li, G. Hu, Q. Lian, and R.S. Ge. 2015. In utero exposure to diisononyl phthalate caused testicular dysgenesis of rat fetal testis. Toxicol. Lett. 232(2):466-474.

Li, M., L. Qiu, Y. Zhang, Y. Hua, S. Tu, Y. He, S. Wen, Q. Wang, and G. Wei. 2013. Dose-related effect by maternal exposure to di-(2-ethylhexyl) phthalate plasticizer on inducing hypospadiac male rats. Environ. Toxicol. Pharmacol. 35(1):55-60.

Li, N., X. Chen, X. Zhou, W. Zhang, J. Yuan, and J. Feng. 2015. The mechanism underlying dibutyl phthalate induced shortened anogenital distance and hypospadias in rats. J. Pediatr. Surg. 50(12):2078-2083.

Li, Y., M. Zhuang, T. Li, and N. Shi. 2009. Neurobehavioral toxicity study of dibutyl phthalate on rats following in utero and lactational exposure. J. Appl. Toxicol. 29(7):603-611.

Lin, H., R.S. Ge, G.R. Chen, G.X. Hu, L. Dong, Q.Q. Lian, D.O. Hardy, C.M. Sottas, X.K. Li, and M.P. Hardy. 2008. Involvement of testicular growth factors in fetal Leydig cell aggregation after exposure to phthalate in utero. Proc. Natl. Acad. Sci. U.S.A. 105(20):7218-7222.

Lin, H., Q.Q. Lian, G.X. Hu, Y. Jin, Y. Zhang, D.O. Hardy, G.R. Chen, Z.Q. Lu, C.M. Sottas, M.P. Hardy, and R.S. Ge. 2009. In utero and lactational exposures to diethylhexyl-phthalate affect two populations of Leydig cells in male Long-Evans rats. Biol. Reprod. 80(5):882-888.

Lin, L.C., S.L. Wang, Y.C. Chang, P.C. Huang, J.T. Cheng, P.H. Su, and P.C. Liao. 2011. Associations between maternal phthalate exposure and cord sex hormones in human infants. Chemosphere 83(8):1192-1199.

Lioy, P.J., R. Hauser, C. Gennings, H.M Koch, P.E. Mirkes, B.A. Schwetz, and A. Kortenkamp. 2015. Assessment of phthalates/phthalate alternatives in children's toys and childcare articles: Review of the report including conclusions and recommendation of the Chronic Hazard Advisory Panel of the Consumer Product Safety Commission. J. Expo. Sci. Environ. Epidemiol. 5(4):343-353.

Liu, X., D.W. He, D.Y. Zhang, T. Lin, and G.H. Wei. 2008. Di(2-ethylhexyl) phthalate (DEHP) increases transforming growth factor-beta1 expression in fetal mouse genital tubercles. J. Toxicol. Environ. Health A 71(19):1289-1294.

Lorber, M., and A.M. Calafat. 2012. Dose reconstruction of di(2-ethylhexyl) phthalate using a simple pharmacokinetic model. Environ. Health Perspect. 120(12):1705-1710.

Lorber, M., J. Angerer, and H.M. Koch. 2010. A simple pharmacokinetic model to characterize exposure of Americans to di-2-ethylhexyl phthalate. J. Expo. Sci. Environ. Epidemiol. 20(1):38-53.

MacLeod, D.J., R.M. Sharpe, M. Welsh, M. Fisken, H.M. Scott, G.R. Hutchison, A.J. Drake, and S. van den Driesche. 2010. Androgen action in the masculinization programming window and development of male reproductive organs. Int. J. Androl. 33(2):279-287.

Mahood, I.K., H.M. Scott, R. Brown, N. Hallmark, M. Walker, and R.M. Sharpe. 2007. In utero exposure to di(nbutyl) phthalate and testicular dysgenesis: Comparison of fetal and adult end points and their dose sensitivity. Environ. Health Perspect. 115(Suppl. 1):55-61.

Martinez-Arguelles, D.B., T. Guichard, M. Culty, B.R. Zirkin, and V. Papadopoulos. 2011. In utero exposure to the antiandrogen di-(2-ethylhexyl) phthalate decreases adrenal aldosterone production in the adult rat. Biol. Reprod. 85(1):51-61.

Martinez-Arguelles, D.B., E. Campioli, M. Culty, B.R. Zirkin, and V. Papadopoulos. 2013. Fetal origin of endocrine dysfunction in the adult: The phthalate model. J. Steroid Biochem. Mol. Biol. 137:5-17.

Martino-Andrade, A.J., R.N. Morais, G.G. Botelho, G. Muller, S.W. Grande, G.B. Carpentieri, G.M. Leao, and P.R. Dalsenter. 2009. Coadministration of active phthalates results in disruption of foetal testicular function in rats. Int. J. Androl. 32(6):704-712.

Martino-Andrade, A.J., F. Liu, S. Sathyanarayana, E.S. Barrett, J.B. Redmon, R.H. Nguyen, H. Levine, and S.H. Swan. 2016. Timing of prenatal phthalate exposure in relation to genital endpoints in male newborns. Andrology 4(4):585-593.

Masutomi, N., M. Shibutani, H. Takagi, C. Uneyama, N. Takahashi, and M. Hirose. 2003. Impact of dietary exposure to methoxychlor, genistein, or diisononyl phthalate during the perinatal period on the development of the rat endocrine/reproductive systems in later life. Toxicology 192(2-3):149-170.

McIntyre, B.S., N.J. Barlow, and P.M. Foster. 2001. Androgen-mediated development in male rat offspring exposed to flutamide in utero: Permanence and correlation of early postnatal changes in anogenital distance and nipple retention with malformations in androgen-dependent tissues. Toxicol. Sci. 62(2):236-249.

McKinnell, C., R.T. Mitchell, M. Walker, K. Morris, C.J. Kelnar, W.H. Wallace, and R.M. Sharpe. 2009. Effect of fetal or neonatal exposure to monobutyl phthalate (MBP) on testicular development and function in the marmoset. Hum. Reprod. 24(9):2244-2254.

Mitchell, R.T., A.J. Childs, R.A. Anderson, S. van den Driesche, P.T. Saunders, C. McKinnell, W.H. Wallace, C.J. Kelnar, and R.M. Sharpe. 2012. Do phthalates affect steroidogenesis by the human fetal testis? Exposure of human fetal testis xenografts to di-n-butyl phthalate. J. Clin. Endocrinol. Metab. 97(3):E341-E348.

Moore, R.W., T.A. Rudy, T.M. Lin, K. Ko, and R.E. Peterson. 2001. Abnormalities of sexual development in male rats with in utero and lactational exposure to the antiandrogenic plasticizer di(2-ethylhexyl) phthalate. Environ. Health Perspect. 109(3):229-237.

Mylchreest, E., R.C. Cattley, and P.M. Foster. 1998. Male reproductive tract malformations in rats following gestational and lactational exposure to di(n-butyl) phthalate: An antiandrogenic mechanism? Toxicol. Sci. 43(1):47-60.

Mylchreest, E., M. Sar, R.C. Cattley, and P.M. Foster. 1999. Disruption of androgen-regulated male reproductive development by di(n-butyl) phthalate during late gestation in rats is different from flutamide. Toxicol. Appl. Pharmacol. 156(2):81-95.

Mylchreest, E., D.G. Wallace, R.C. Cattley, and P.M. Foster. 2000. Dose-dependent alterations in androgen-regulated male reproductive development in rats exposed to di(n-butyl) phthalate during late gestation. Toxicol. Sci. 55(1):143-151.

Nagao, T., R. Ohta, H. Marumo, T. Shindo, S. Yoshimura, and H. Ono. 2000. Effect of butyl benzyl phthalate in Sprague-Dawley rats after gavage administration: A two-generation reproductive study. Reprod. Toxicol. 14(6):513-532.

NRC (National Research Council). 2006. Assessing the Human Health Effects of Trichloroethylene. Washington, DC: The National Academies Press.

NRC. 2008. Phthalates and Cumulative Risk Assessment: The Tasks Ahead. Washington, DC: The National Academies Press.

NTP (National Toxicology Program). 2015. Handbook for Conducting a Literature-Base Health Assessment Using OHAT Approach for Systematic Review and Evidence Integration, January 9, 2015. Office of Health Assessment and Translation, Division of the National Toxicology Program, National Institute of Environmental Health Sciences [online]. Available: https://ntp.niehs.nih.gov/ntp/ohat/pubs/handbookjan2015_508.pdf [accessed June 17, 2016].

Pocar, P., N. Fiandanese, C. Secchi, A. Berrini, B. Fischer, J.S. Schmidt, K. Schaedlich, and V. Borromeo. 2012. Exposure to di(2-ethyl-hexyl) phthalate (DEHP) in utero and during lactation causes long-term pituitary-gonadal axis disruption in male and female mouse offspring. Endocrinology 153(2):937-948.

Raudenbush, S.W. 2009. Analyzing effect sizes: Random-effects models. Pp. 295-315 in The Handbook of Research Synthesis and Meta-Analysis, 2nd Ed., H. Cooper, L.V. Hedges, and J.C. Valentine, eds. New York: Russell Sage Foundation.

Rooney, A.A., A.L. Boyles, M.S. Wolfe, J.R. Bucher, and K.A. Thayer. 2014. Systematic review and evidence integration for literature-based environmental health assessments. Environ. Health Perspect. 122(7):711-718.

Saillenfait, A.M., J.P. Payan, J.P. Fabry, D. Beydon, I. Langonne, F. Gallissot, and J.P. Sabate. 1998. Assessment of the developmental toxicity, metabolism, and placental transfer of di-n-butyl phthalate administered to pregnant rats. Toxicol. Sci. 45(2):212-224.

Saillenfait, A.M., J.P. Sabate, and F. Gallissot. 2008. Diisobutyl phthalate impairs the androgen-dependent reproductive development of the male rat. Reprod. Toxicol. 26(2):107-115.

Saillenfait, A.M., J.P. Sabate, and F. Gallissot. 2009. Effects of in utero exposure to di-n-hexyl phthalate on the reproductive development of the male rat. Reprod. Toxicol. 28(4):468-476.

Saillenfait, A.M., A.C. Roudot, F. Gallissot, and J.P. Sabate. 2011. Prenatal developmental toxicity studies on di-nheptyl and di-n-octyl phthalates in Sprague-Dawley rats. Reprod. Toxicol. 32(3):268-276.

Saillenfait, A.M., J.P. Sabate, A. Robert, B. Cossec, A.C. Roudot, F. Denis, and M. Burgart. 2013a. Adverse effects of diisooctyl phthalate on the male rat reproductive development following prenatal exposure. Reprod. Toxicol. 42:192-202.

Saillenfait, A.M., J.P. Sabate, A. Robert, V. Rouiller-Fabre, A.C. Roudot, D. Moison, and F. Denis. 2013b. Dose-dependent alterations in gene expression and testosterone production in fetal rat testis after exposure to di-n-hexyl phthalate. J. Appl. Toxicol. 33(9):1027-1035.

Samandar, E., M.J. Silva, J.A. Reidy, L.L Needham, and A.M. Calafat. 2009. Temporal stability of eight phthalate metabolites and their glucuronide conjugates in human urine. Environ. Res. 109(5):641-646.

Sathyanarayana, S., E. Barrett, S. Butts, C. Wang, and S.H. Swan. 2014. Phthalate exposure and reproductive hormone concentrations in pregnancy. Reproduction 147(4):401-409.

Scarano, W.R., F.C. Toledo, M.T. Guerra, P.F. Pinheiro, R.F. Domeniconi, S.L. Felisbino, S.G. Campos, S.R. Taboga, and W.G. Kempinas. 2010. Functional and morphological reproductive aspects in male rats exposed to di-n-butyl phthalate (DBP) in utero and during lactation. J. Toxicol. Environ. Health A 73(13-14):972-984.

Scott, H.M., J.I. Mason, and R.M. Sharpe. 2009. Steroidogenesis in the fetal testis and its susceptibility to disruption by exogenous compounds. Endocr. Rev. 30(7):883-925.

Sena, E.S., G.L. Currie, S.K. McCann, M.R. Macleod, and D.W. Howells. 2014. Systematic reviews and meta-analysis of preclinical studies: why perform them and how to appraise them critically. J. Cereb. Blood Flow Metab. 34(5):737-742.

Silva, M.J., J.A. Reidy, A.R. Herbert, J.L. Preau, L.L. Needham, and A.M. Calafat. 2004. Detection of phthalate metabolites in human aminiotic fluid. Bull. Environ. Contam. Toxicol. 72(6):1226-1231.

Skakkebaek, N.E. 2002. Endocrine disrupters and testicular dysgenesis syndrome. Horm. Res. 57(Suppl. 2):43.

Spade, D.J., E.V. McDonnell, N.E. Heger, J.A. Sanders, C.M. Saffarini, P.A. Gruppuso, M.E. De Paepe, and K. Boekelheid. 2014. Xenotransplantation models to study the effects of toxicants on human fetal tissues. Birth Defects Res. B Dev. Reprod. Toxicol. 101(6):410-422.

Stroheker, T., J.F. Regnier, J. Lassurguere, and M.C. Chagnon. 2006. Effect of in utero exposure to di-(2ethylhexyl)phthalate: Distribution in the rat fetus and testosterone production by rat fetal testis in culture. Food Chem. Toxicol. 44(12):2064-2069.

Struve, M.F., K.W. Gaido, J.B. Hensley, K.P. Lehmann, S.M. Ross, M.A. Sochaski, G.A. Willson, and D.C. Dorman. 2009. Reproductive toxicity and pharmacokinetics of di-n-butyl phthalate (DBP) following dietary exposure of pregnant rats. Birth Defects Res. B Dev. Reprod. Toxicol. 86(4):345-354.

Sullivan, S., and J.L. Teague. 2005. Decrease in anogenital distance among male infants with prenatal phthalate exposure. Environ. Health Perspect. 113(8):1056-1061.

Suzuki, Y., J. Yoshinaga, Y. Mizumoto, S. Serizawa, and H. Shiraishi. 2012. Foetal exposure to phthalate esters and anogenital distance in male newborns. Int. J. Androl. 35(3):236-244.

Swan, S.H. 2008. Environmental phthalate exposure in relation to reproductive outcomes and other health endpoints in humans. Environ. Res. 108(2):177-184.

Swan, S.H., K.M. Main, F. Liu, S.L. Stewart, R.L. Kruse, A.M. Calafat, C.S. Mao, J.B. Redmon, C.L. Ternand, S. Sullivan, and J.L. Teague. 2005. Decrease in anogenital distance among male infants with prenatal phthalate exposure. Environ. Health Perspect. 113(8):1056-1061.

Swan, S.H., S. Sathyanarayana, E.S. Barrett, S. Janssen, F. Liu, R.H. Nguyen, and J.B. Redmon. 2015. First trimester phthalate exposure and anogenital distance in newborns. Hum. Reprod. 30(4):963-972.

Thankamony, A., N. Lek, D. Carroll, M. Williams, D.B. Dunger, C.L. Acerini, K.K. Ong, and I.A. Hughes. 2014. Anogenital distance and penile length in infants with hypospadias or cryptorchidism: Comparison with normative data. Environ. Health Perspect. 122(2):207-211.

Thompson, C.J., S.M. Ross, J. Hensley, K. Liu, S.C. Heinze, S.S. Young, and K.W. Gaido. 2005. Differential steroidogenic gene expression in the fetal adrenal gland versus the testis and rapid and dynamic response of the fetal testis to di(n-butyl) phthalate. Biol. Reprod. 73(5):908-917.

Tyl, R.W., C.B. Myers, M.C. Marr, P.A. Fail, J.C. Seely, D.R. Brine, R.A. Barter, and J.H. Butala. 2004. Reproductive toxicity evaluation of dietary butyl benzyl phthalate (BBP) in rats. Reprod. Toxicol. 18(2):241-264.

van den Driesche, S., P. Kolovos, S. Platts, A.J. Drake, and R.M. Sharpe. 2012. Inter-relationship between testicular dysgenesis and Leydig cell function in the masculinization programming window in the rat. PLOS One 7(1):e30111.

van der Zanden, L.F., I.A. van Rooij, W.F. Feitz, B. Franke, N.V. Knoers, and N. Roeleveld. 2012. Aetiology of hypospadias: A systematic review of genes and environment. Hum. Reprod. Update 18(3):260-283.

Viechtbauer, W. 2005. Bias and efficiency of meta-analytic variance estimators in the random-effects model. J. Educ. Behavior. Stat. 30(3):261-293.

Viechtbauer, W. 2010. Conducting meta-analysis in R with the metaphor package. J. Stat. Softw.36(3):1-48.

Vo, T.T., E.M. Jung, V.H. Dang, K. Jung, J. Baek, K.C. Choi, and E.B. Jeung. 2009. Differential effects of flutamide and di-(2-ethylhexyl) phthalate on male reproductive organs in a rat model. J. Reprod. Dev. 55(4):400-411.

Welsh, M., P.T. Saunders, M. Fisken, H.M. Scott, G.R. Hutchison, L.B. Smith, and R.M. Sharpe. 2008. Identification in rats of a programming window for reproductive tract masculinization, disruption of which leads to hypospadias and cryptorchidism. J. Clin. Invest. 118(4):1479-1490.

Wilson, V.S., K.L. Howdeshell, C.S. Lambright, J. Furr, and L.E. Gray, Jr. 2007. Differential expression of the phthalate syndrome in male Sprague-Dawley and Wistar rats after in utero DEHP exposure. Toxicol. Lett. 170(3):177-184.

Wilson, V.S., C.R. Blystone, A.K. Hotchkiss, C.V. Rider, and L.E. Gray, Jr. 2008. Diverse mechanisms of antiandrogen action: Impact on male rat reproductive tract development. Int. J. Androl. 31(2):178-187.

Wittassek, M., and J. Angerer. 2008. Phthalates: Metabolism and exposure. Int. J. Androl. 31(2):131-138.

Wittassek, M., J. Angerer, M. Kolossa-Gehring, S.D. Schafer, W. Klockenbusch, L. Dober, A.K. Gunsel, A. Muller, and G.A. Wiesmuller. 2009. Fetal exposure to phthalates—a pilot study. Int. J. Hyg. Environ. Health 212(5):492-498.

Wohlfahrt-Veje, C., K.M. Main, and N.E. Skakkebaek. 2009. Testicular dysgenesis syndrome: Foetal origin of adult reproductive problems. Clin. Endocrinol. 71(4):459-465.

Wolfe, G.W., and K.A. Layton. 2005. Diethylhexylphthalate: Multigenerational Reproductive Assessment by Continuous Breeding when Administered to Sprague-Dawley Rats in the Diet. TRC Study No. 7244-200. NTPRACB 98-004. TherImmune Research Corporation, Gaithersburg, MD.

Wolfe, G.W., and R.V. Patel. 2002. Dibutyl Phthalate: Multigenerational Reproductive Assessment by Continuous Breeding When Administered to Sprague-Dawley Rats in the Diet, Vol.1. NTP RACB 97003. TherImmune Research Corporation, Gaithersburg, MD.

Zhang, L.D., Q. Deng, Z.M. Wang, M. Gao, L. Wang, T. Chong, and H.C. Li. 2013. Disruption of reproductive development in male rat offspring following gestational and lactational exposure to di-(2-ethylhexyl) phthalate and genistein. Biol. Res. 46(2):139-146.

Zhang, Y., X. Jiang, and B. Chen. 2004. Reproductive and developmental toxicity in F1 Sprague-Dawley male rats exposed to di-n-butyl phthalate in utero and during lactation and determination of its NOAEL. Reprod. Toxicol. 18(5):669-676.

Zota, A.R., A.M. Calafat, and T.J. Woodruff. 2014. Temporal trends in phthalate exposure: Findings from the National Health and Nutrition Examination Survey, 2001-2010. Environ. Health Perspect. 122(3):235-241.