GESTATIONAL PHTHALATE/REPLACEMENT EXPOSURE: A GLIMPSE INTO MATERNAL RISK FACTORS, BIOLOGICAL TARGETS, AND GESTATIONAL CARDIOMETABOLIC HEALTH By Diana C. Pacyga A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of Human Nutrition – Environmental Toxicology – Doctor of Philosophy 2023 ABSTRACT Pregnancy is a period of heightened susceptibility to numerous modifiable environmental stressors, especially to one diverse class of endocrine and metabolic disrupting chemicals, ortho-phthalate diesters (commonly known as phthalates) and their replacements. Evidence suggests prenatal phthalate exposure is associated with adverse maternal and child health outcomes in pregnancy, but these health consequences may also persist well beyond gestation. This is concerning, as pregnant women are ubiquitously exposed to phthalates, but also their plasticizer replacements di(isononyl) cyclohexane-1,2-dicarboxylate (DiNCH) and di(2-ethylhexyl) terephthalate (DEHTP) to which exposure is increasing. Based on recent experimental and observational evidence, these replacements may have similar or worse health consequences than the original phthalates. Additionally, pregnant women are not exposed to one phthalate or replacement at a time, but a mixture of these chemicals, which necessitates evaluating associations of multiple phthalates/replacements with our outcomes of interest to understand the potential true impact of real-life exposures. Therefore, the studies presented in this dissertation were designed to identify potential maternal risk factors, gestational hormonal targets, and cardiometabolic health consequences of prenatal exposure to phthalates and their replacements in women enrolled in the Illinois Kids Development Study. Specifically, we evaluated determinants of maternal phthalate/ replacement exposure (Chapter 2), as well as associations of phthalates/replacements with maternal sex-steroid hormones (Chapter 3) and gestational weight gain (Chapter 4). We developed an approach for measuring urinary sex-steroid hormones at multiple gestational timepoints to capture longitudinal changes in associations of phthalates/ replacements with hormones. Given the roles of gestational hormones in coordinating critical pregnancy metabolic adaptations, we also addressed the potential involvement of these biological processes in associations between phthalates/replacements and maternal cardiometabolic health, with a focus on gestational weight gain as one clinically- relevant metabolic endpoint. Throughout, various statistical mixtures approaches, including weighted quantile sum regression, quantile-based g-computation, k-means clustering, and principal component analysis, were highlighted as ways to evaluate phthalate/replacement mixtures. The overarching goal of this dissertation was to underscore the importance of considering maternal pregnancy health, highlighting regrettable substitution, and emphasizing the utility of statistical mixtures approaches to address our research questions of interest. Copyright by DIANA C. PACYGA 2023 ACKNOWLEDGEMENTS First, I would like to thank my advisor, Dr. Rita Strakovsky, for her encouragement and guidance over the past decade. I am extremely grateful for the supportive environment and amazing opportunities she provided me, one of which resulted in me pursuing graduate school and discovering my love for science. I will forever be thankful for her commitment to and belief in me. I am also grateful to my committee members, Drs. Joseph Gardiner, Sarah Comstock, Felicia Wu, and Nicole Talge. It was wonderful to have their support as I navigated my program. I am proud to say that I got to know and work closely with each and every one of my committee members as they shared their expertise in biostatistics, nutrition, toxicology, and epidemiology with me. Special thanks to Drs. Susan Schantz, Jodi Flaws, Diana Haggerty, Joseph Braun, George Papandonatos, Rebecca Smith, Paige Williams, Antonia Calafat, Tamarra James-Todd, and Susan Korrick. I am extremely fortunate to have a great group of collaborators who are very generous with their time and expertise in perinatal epidemiology, women’s health, environmental health, toxicology, and complex biostatistical methods. The work in my dissertation would not be possible without the Illinois Kids Development Study (I-KIDS) team and participants. I started my career as an I-KIDS undergraduate research assistant and saw first-hand how much time the team of volunteers, students, and support staff donated to make I-KIDS successful. I am also very appreciative of the I-KIDS families who enrolled in the study – their dedication and commitment to science is truly amazing. v None of my accomplishments would be possible without the departmental and program support at Michigan State University. Thank you to the Department of Food Science & Human Nutrition and the Environmental & Integrative Toxicological Sciences Program for providing a supportive environment, generous scholarships, and academic training that are invaluable for my professional development. I am also grateful for the wonderful community of fellow trainees. I have so gained many friendships that I will cherish forever. Last, but most importantly, I could not have accomplished any of this without the support and encouragement from my friends and family. To my loving parents and brother who are my biggest cheerleaders. I will always be grateful for the hard work my parents put in over the years to provide me with everything I needed to be successful in my pursuits. It is because of them I learned the importance of taking advantage of and valuing every opportunity that comes my way. They are my inspiration. To my wonderful friends and family in Michigan, Idaho, Illinois, and Poland who cheered me on from near and afar. Thank you so much for the support and encouragement. “Najpiękniejszych chwil w życiu nie zaplanujesz. One przyjdą same.” vi TABLE OF CONTENTS LIST OF ABBREVIATIONS ........................................................................................... viii CHAPTER 1: INTRODUCTION .................................................................................... 1 CHAPTER 2: IDENTIFICATION OF PROFILES AND DETERMINANTS OF MATERNAL PREGNANCY URINARY BIOMARKERS OF PHTHALATES AND REPLACEMENTS IN THE ILLINOIS KIDS DEVELOPMENT STUDY ........................................................................ 9 CHAPTER 3: MATERNAL PHTHALATE AND PHTHALATE ALTERNATIVE METABOLITES AND URINARY BIOMARKERS OF ESTROGENS AND TESTOSTERONES ACROSS PREGNANCY ............................... 48 CHAPTER 4: ASSOCIATIONS OF INDIVIDUAL AND CUMULATIVE URINARY PHTHALATE AND REPLACEMENT BIOMARKERS WITH GESTATIONAL WEIGHT GAIN THROUGH LATE PREGNANCY ........ 87 CHAPTER 5: CONCLUSIONS AND FUTURE DIRECTIONS ................................... 130 BIBLIOGRAPHY ......................................................................................................... 146 APPENDIX A: DIETARY PREDICTORS OF PHTHALATE AND BISPHENOL EXPOSURES IN PREGNANT WOMEN .............................................. 214 APPENDIX B: MATERNAL DIET QUALITY MODERATES ASSOCIATIONS BETWEEN PARABENS AND BIRTH OUTCOMES............................. 239 APPENDIX C: ASSOCIATIONS OF PREGNANCY HISTORY WITH BMI AND WEIGHT GAIN IN 45-54-YEAR-OLD WOMEN ................................... 274 APPENDIX D: URINARY PHTHALATE METABOLITE CONCENTRATIONS AND SERUM HORMONE LEVELS IN PRE- AND PERIMENOPAUSAL WOMEN FROM THE MIDLIFE WOMEN’S HEALTH STUDY ............. 303 APPENDIX E: URINARY PHTHALATE METABOLITE CONCENTRATIONS AND HOT FLASHES IN WOMEN FROM AN URBAN CONVENIENCE SAMPLE OF MIDLIFE WOMEN .......................................................... 338 APPENDIX F: MIDLIFE URINARY PHTHALATE METABOLITE CONCENTRATIONS AND PRIOR UTERINE FIBROID DIAGNOSIS ......................................................................................... 370 vii LIST OF ABBREVIATIONS AA Anti-androgenic AHEI-2010 Alternative Healthy Eating Index AMH anti-Müllerian hormone BBzP Benzylbutyl phthalate BKMR Bayesian Kernel Machine Regression BMI Body mass index BPA Bisphenol A BPF Bisphenol F BPS Bisphenol S CDC Centers for Disease Control and Prevention CESD Center for Epidemiological Studies Depression CI Confidence interval CVD Cardiovascular Disease DAG Directed acyclic graph DBP Di-n-butyl phthalate DEHTP Di(2-ethylhexyl) terephthalate DEP Diethyl phthalate DiBP Di-iso-butyl phthalate DiNCH Di(isononyl) cyclohexane-1,2-dicarboxylate DiNP Di-isononyl phthalate DOP Di-n-octyl phthalate EDC Endocrine disrupting chemical viii EDPS Edinburgh Postnatal Depression Scale ELISA Enzyme-linked immunosorbent assay FFQ Food Frequency Questionnaire FSH Follicle-stimulating hormone GWG Gestational weight gain GWGz Gestational weight gain z-scores HDL High-density lipoprotein HighMWP High-molecular-weight phthalate HPG Hypothalamic-pituitary-gonadal HPLC-MS/MS High-performance liquid chromatography mass spectrometry I-KIDS Illinois Kids Development Study IOM Institute of Medicine IQR Interquartile range LDL Low-density lipoprotein LH Luteinizing hormone LOD Limit of detection LOQ Limit of quantification LowMWP Low-molecular-weight phthalate MBP Mono-n-butyl phthalate MBzP Monobenzyl phthalate MCNP Monocarboxynonyl phthalate MCOCH Cyclohexane-1,2-dicarboxylic acid-mono(carboxyoctyl) ester MCOP Monocarboxyoctyl phthalate ix MCPP Mono(3-carboxypropyl) phthalate MECPP Mono(2-ethyl-5-carboxypentyl) phthalate MECPTP Mono(2-ethyl-5-carboxypentyl) terephthalate MEHP Mono(2-ethylhexyl) phthalate MEHHP Mono(2-ethyl-5-hydroxyhexyl) phthalate MEHHTP Mono(2-ethyl-5-hydroxyhexyl) terephthalate MEOHP Mono(2-ethyl-5-oxohexyl) phthalate MEP Monoethyl phthalate MHBP Mono-hydroxybutyl phthalate MHiBP Mono-hydroxy-isobutyl phthalate MHiNCH Cyclohexane-1,2-dicarboxylic acid-monohydroxy isononyl ester MiBP Mono-isobutyl phthalate MiNP Mono-isononyl phthalate MONP Monooxononyl phthalate MRI Magnetic resonance imaging MWHS Midlife Women’s Health Study NHANES National Health and Nutrition Examination Survey OR Odds Ratio PC Principal component PCA Principal component analysis PCP Personal care products Pint P-value for moderation tested using multiplicative interactions Plinear trend P-value for linear trend test across quartiles x Q1 Quartile 1 Q2 Quartile 2 Q3 Quartile 3 Q4 Quartile 4 QGComp Quantile-based G-computation RIA Radioimmunoassay RR Risk ratio SES Socioeconomic status SG Specific gravity SHBG Sex hormone binding globulin STRAW+10 Stages of Reproductive Aging Workshop + 10 ∑AA Sum of phthalate metabolites with anti-androgenic biological activity ∑DBP Sum of di-n-butyl phthalate metabolites ∑DEHP Sum of di(2-ethylhexyl) phthalate metabolites ∑DEHTP Sum of di(2-ethylhexyl) terephthalate metabolites ∑DiBP Sum of di-iso-butyl phthalate metabolites ∑DiNCH Sum of di(isononyl) cyclohexane-1,2-dicarboxylate metabolites ∑DiNP Sum of di-isononyl phthalate metabolites ∑PCP Sum of phthalate metabolites of parents in personal care products ∑Phthalates Sum of all phthalate metabolites ∑Plastics Sum of phthalate metabolites of parents in plastics SumDBP Sum of di-n-butyl phthalate metabolites SumDEHP Sum of di(2-ethylhexyl) phthalate metabolites xi SumDEHTP Sum of di(2-ethylhexyl) terephthalate metabolites SumDiBP Sum of di-iso-butyl phthalate metabolites SumDiNCH Sum of di(isononyl) cyclohexane-1,2-dicarboxylate metabolites SumDiNP Sum of di-isononyl phthalate metabolites SumEstrogens Sum of the eight major urinary estrogen metabolites SumPCP Sum of phthalate metabolites of parents in personal care products SumPlastics Sum of phthalate metabolites of parents in plastics SumTestosterones Sum of the two major testosterone metabolites TDS Total Diet Study U.S. United States VLDL Very-low-density lipoprotein WQSR Weighted Quantile Sum Regression xii CHAPTER 1: INTRODUCTION As many as 86% of women in the United States (U.S.) will have given birth at least once by the time they reach their forties (1, 2). Pregnancy is a time of positive energy balance requiring weight gain to support not only the growing fetus and placenta, but also to increase maternal tissue deposition and plasma volume for nutrient storage and fetal transfer (3). These processes are coordinated via physiological adaptations in hormone synthesis and signaling (4), in glucose and lipid homeostasis (3), and in inflammation and immune response (5). For example, substantial increases in estrogens (i.e., estrone, estradiol, and estriol) and minor increases in androgens (i.e., testosterone) support implantation, placental angiogenesis, parturition, and metabolic adaptations (3, 4, 6, 7). However, this also makes pregnancy a sensitive window since disruptions in the above- mentioned adaptations can lead to adverse pregnancy and birth outcomes, including pre- term birth, altered fetal growth, pregnancy hypertensive disorders, and gestational diabetes (7-9). Consequently, perturbations during pregnancy can have lasting consequences for child health (2, 10, 11). For example, adverse pregnancy and birth outcomes have been associated with the development of later life diseases in the offspring, including cardiovascular disease, metabolic syndrome, osteoporosis, infertility, and cognitive dysfunction (10). This is in accordance with the developmental origins of health and disease hypothesis (10, 12). Pregnancy is also an important determinant of maternal lifelong health, as more recent preliminary evidence indicates that pregnancy itself (and especially having adverse pregnancy or birth outcomes) may have lasting consequences for maternal later life cardiometabolic disease risk (2, 11). Therefore, it is important to identify the modifiable risk factors associated with adverse pregnancy health 1 to protect maternal and child lifelong health. Pregnancy is a period of heightened susceptibility to numerous modifiable environmental stressors (2, 13), especially to one diverse class of endocrine and metabolic disrupting chemicals, ortho-phthalate diesters (commonly known as phthalates) and their replacements. Greater than 90% of pregnant women have measurable levels of urinary phthalate metabolites, which are used as biomarkers for predicting exposure to phthalates, suggesting that pregnant women are ubiquitously exposed to these chemicals (14). This widespread exposure is due to the use of phthalates in common daily used consumer products. For example, some phthalates, such as di(2-ethylhexyl) phthalate (DEHP), are plasticizers used during food processing and in food packaging materials, but can also be found in medical tubing and in the coating of certain medications and supplements (15). Other phthalates, such as diethyl phthalate (DEP), are solvents used to stabilize scents in personal care products and cosmetics (16). As a result, ingestion, dermal absorption, and inhalation are major routes of exposure to these chemicals during pregnancy. Because these chemicals are found in personal care products and cosmetics, women generally have higher urinary phthalate metabolite concentrations than men (14). Additionally, a pregnant woman’s phthalate exposure largely depends on her socioeconomic status, lifestyle, and health factors (17, 18). This is important to note since knowing the major risk factors associated with higher or even lower phthalate exposure is not only necessary for understanding exposure patterns, but also critical for identifying at risk groups of pregnant women. The current literature suggests that pregnant women who self-identify as Black and/or Hispanic, have lower socioeconomic status, and/or 2 engage in unhealthy lifestyle behaviors are likely to have higher phthalate exposure than non-Hispanic White women, those with higher socioeconomic status, and/or those who engage in healthier lifestyle behaviors (17, 18). Further research is necessary to corroborate these findings, but also identify other critical determinants, particularly modifiable factors, of phthalate exposure that may be exclusive to pregnancy. Phthalates are traditionally classified as endocrine and metabolic disrupting chemicals based on evidence from experimental models in non-pregnant animals, although there are limited and inconsistent findings in pregnancy experimental and observational studies. Specifically, in vitro studies suggest that phthalates are weakly estrogenic (19, 20), and in vivo studies in male offspring of exposed dams showed that certain phthalates have anti-androgenic effects (21, 22). The metabolic disrupting properties are supported by studies in vitro and non-pregnancy in vivo models showing that phthalates can interact with peroxisome proliferator-activated receptor gamma, liver X receptors, and retinoid X receptors (23) – these receptors are important mediators for regulating glucose and lipid homeostasis, adipogenesis, but also hormone synthesis and signaling. Because pregnancy is a sensitive window with coordinated dynamic changes in hormone and metabolic homeostasis, the endocrine and metabolic disrupting potential of phthalates and widespread exposure to these chemicals is concerning. Unfortunately, it is also these pregnancy-specific physiological adaptations that makes it challenging to translate the above mentioned experimental findings to human pregnant women. Pregnancy adaptations are also trimester-specific, making it challenging for human epidemiologic studies to identify consistent associations across various cohorts. Studies in pregnancy 3 cohorts are needed to determine if the endocrine and metabolic properties of phthalates translate to human pregnant women. The above-discussed phthalate mechanisms of action are especially detrimental in pregnancy since they may be responsible for adverse pregnancy and birth outcomes, including altered gestational weight gain – a clinically relevant marker of gestational health outcomes. In pregnancy cohort studies, prenatal phthalate exposure is associated with increased odds of pregnancy loss (24), pregnancy hypertension (25, 26), gestational diabetes (27), pre-term birth (28-30), altered birth weight (26), and inappropriate gestational weight gain (31-33). The deleterious effects of phthalates on pre-term birth, altered birth weight, and inappropriate gestational weight gain has also been reported in several in vivo pregnancy experimental animal models (34-39). Clinically, gestational weight gain is used as an easily measured marker of pregnancy health and fetal growth because it is a complex phenotype with contributions from both the mother and the developing fetus (3). Institute of Medicine (IOM) recommendations for pregnancy weight gain were implemented based on the robust evidence that deviations from appropriate weight gain are predictive of the same above-mentioned poor pregnancy and birth outcomes (40-42). Inappropriate weight gain may also have long lasting consequences for both mother and baby (43, 44), which are also potential adverse endpoints of phthalate exposure. Conducting further studies to determine whether phthalates are associated with altered gestational weight gain will be important to elucidate the potential maternal and child health repercussions of prenatal phthalate exposure. 4 In response to growing public concerns over prenatal phthalate exposure for child health and the subsequent regulation of these chemicals in certain consumer products (58), purportedly safe replacements were introduced. Specifically, replacements di(isononyl) cyclohexane-1,2-dicarboxylate (DiNCH) and a terephthalate diester, di(2-ethylhexyl) terephthalate (DEHTP) were developed and introduced into the U.S. market before or in early 2000s to replace plasticizer phthalates such as DEHP (59-61). Unfortunately, the safety and toxicity of these ortho-phthalate replacements are relatively unknown, but recent observational evidence indicates that prenatal exposure to DiNCH and DEHTP may be associated with increased risk of pre-term birth (62) and inappropriate gestational weight gain (54). This highlights the concept of regrettable substitution where a chemical with unknown or unforeseen hazard is used to replace a chemical identified as problematic (63). Unfortunately, findings from the National Health and Nutrition Survey (NHANES) and other U.S. biomonitoring studies show that urinary ortho-phthalate metabolite concentrations are decreasing, while DiNCH and DEHTP metabolite concentrations are increasing over time (64, 65). Additional studies in pregnancy are needed to understand the implications of prenatal DiNCH and DEHTP exposure. Pregnant women are not exposed to one phthalate or replacement at a time, but a mixture of these chemicals. Unfortunately, most prior studies have focused on assessing associations of single phthalates/replacements with health outcomes, which makes it challenging to know the true impact of exposures to real-life chemical mixtures (66). Some studies suggest that the identified risks related to chemical exposures may be greater when evaluating multiple chemicals together rather than one at a time (67). Additionally, 5 assessing chemicals as mixtures may better identify pregnant women who will benefit most from interventions targeted at reducing chemical exposures. Therefore, to better simulate real life exposure, the environmental epidemiology field has turned to utilizing traditional, but also developing novel methods for evaluating chemicals as mixtures (66, 67). For example, traditional unsupervised (pattern identification methods, including clustering (i.e., k-means clustering) and dimension reduction (i.e., principal component analysis, PCA) approaches can be used to identify sub-groups of pregnant women based on their chemical exposures or identify patterns among chemicals that track together, respectively (68, 69). Novel supervised statistical mixtures methods, including weighted quantile sum regression (WQSR) and quantile-based g-computation (QGComp), can be used to estimate the cumulative or joint effect of multiple chemicals on an outcome of interest, but also identify “bad actors”, or which chemicals contribute most to the overall chemical mixture effect (66, 70-73). Studies focusing on assessing associations of phthalates and their replacements as a mixture are needed to further understand gestational exposure patterns to these chemicals, but also better model the potential true effect of phthalate/replacement exposure on pregnancy health. Given the gaps outlined above, the main objectives of this dissertation were to further understand the maternal risk factors, pregnancy hormonal targets, and cardiometabolic consequences of exposure to individual and a mixture phthalates and their replacements in pregnancy. Specifically, we identified the major maternal characteristics associated with prenatal phthalate/replacement exposure and determined if phthalate/replacement exposure was associated with altered gestational hormone concentrations and 6 gestational weight gain. These knowledge gaps were addressed in three aims corresponding to Chapters 2 – 4 of this dissertation (presented in Figure 1) using data collected as part of the Illinois Kids Development Study (I-KIDS) – an ongoing prospective pregnancy and birth cohort of women from Champaign-Urbana, Illinois following 720 pregnant women and their child from early pregnancy through childhood. In Aim 1 (Chapter 2), we used pattern identification methods, k-means and PCA, to identify phthalate/replacement exposure patterns along with major seasonal, time, and maternal sociodemographic, lifestyle, and health determinants of these exposure patterns. In Aim 2 (Chapter 3), we evaluated gestational timepoint- and fetal sex-specific associations of individual phthalates/replacements with maternal estrogen and testosterone concentrations. In Aim 3 (Chapter 4), we evaluated phthalates/replacements individually and as a mixture using WQSR and QGComp to determine if these chemicals are associated with altered gestational weight gain. The overarching goal of this dissertation was to stress the importance of considering maternal pregnancy health, highlight regrettable chemical substitution, and emphasize the utility of statistical mixtures approaches for addressing these research questions. 7 Figure 1. Summary of three aims evaluated in dissertation. The colors represent the following dissertation aims/chapters: blue = Aim 1/Chapter 2, pink = Aim 2/Chapter 3, and purple = Aim 3/Chapter 4. The darker shaded arrows indicate the specific relationships evaluated, while the lighter arrows show how the dissertation aims are related. 8 CHAPTER 2: IDENTIFICATION OF PROFILES AND DETERMINANTS OF MATERNAL PREGNANCY URINARY BIOMARKERS OF PHTHALATES AND REPLACEMENTS IN THE ILLINOIS KIDS DEVELOPMENT STUDY This article/chapter has been published in Environment International; Volume 162; Pacyga DC* and Haggerty DK*, Nicol M, Henning M, Calafat AM, Braun J, Schantz SL, Strakovsky RS, *These authors contributed equally; Identification of profiles and determinants of maternal pregnancy urinary biomarkers of phthalate and phthalate replacements in the Illinois Kids Development Study. Copyright Elsevier (2022); https://doi.org/10.1016/j.envint.2022.107150. 2.1. ABSTRACT Pregnant women are exposed to multiple phthalates and their replacements, which are endocrine disrupting chemicals associated with adverse maternal and child health outcomes. Identifying maternal characteristics associated with phthalate/replacement exposure during pregnancy is important. We evaluated 13 maternal sociodemographic and lifestyle factors, enrollment year, and conception season as determinants of exposure biomarkers of phthalates and their replacements in 482 pregnant women from the Illinois Kids Development Study (I-KIDS, enrolled 2013-2018). We quantified 19 phthalate/replacement metabolites in pools of five first-morning urines collected across pregnancy. K-means clustering identified women with distinct patterns of biomarker concentrations and principal component analysis (PCA) identified principal component (PC) profiles of biomarkers that exist together. We used multivariable regression models to evaluate associations of predictors with identified k-means clusters and PCs. K-means clustering identified two clusters of women: 1) low phthalate/di(2-ethylhexyl) terephthalate 9 (∑DEHTP) and 2) high phthalate/∑DEHTP biomarker concentrations. PCA identified four PCs with loadings heaviest for biomarkers of plasticizer phthalates [di-isononyl, di- isodecyl, di-n-octyl phthalates] (PC1), of other phthalates [dibenzyl, di-n-butyl, di-iso-butyl phthalates] (PC2), of phthalate replacements [∑DEHTP, di(isononyl) cyclohexane-1,2- dicarboxylate (∑DiNCH)] (PC3), and of monoethyl phthalate [MEP] (PC4). Overall, age, marital status, income, parity, pre-pregnancy BMI, caffeine intake, enrollment year, and conception season were independently associated with k-means cluster membership and at least one PC. Additionally, race/ethnicity, education, employment, pregnancy intention, smoking status, alcohol intake, and diet were associated with at least one PC. For instance, women who conceived in the spring, summer, and/or fall months had lower odds of high phthalate/∑DEHTP cluster membership and had lower plasticizer phthalate, phthalate replacement, and MEP PC scores. Conception season, enrollment year, and several sociodemographic/lifestyle factors were predictive of phthalate/replacement biomarker profiles. Future studies should corroborate these findings, with a special focus on replacements to which pregnant women are becoming increasingly exposed. 2.2. KEYWORDS DEHTP, determinants, DiNCH, endocrine disruptors, phthalate, pregnancy. 2.3. INTRODUCTION Ortho-phthalate diesters, or phthalates, are a class of chemicals widely used in the production of plastics for food contact materials and in some personal care products to which humans are exposed through ingestion, dermal absorption, and inhalation (14, 74). 10 An increasing body of evidence points to phthalates as endocrine disrupting chemicals that interact with multiple hormones and hormone-regulated processes (75-79), which is concerning given that pregnancy is a hormonally sensitive window and pregnant women are ubiquitously exposed to these chemicals (14). Higher maternal urinary phthalate metabolite concentrations have been associated with adverse pregnancy outcomes, including preeclampsia and pregnancy hypertensive disorders (25, 26), glucose intolerance and gestational diabetes (32, 80, 81), and preterm birth (82). In response to the growing concerns over the endocrine disrupting properties and subsequent regulation of ortho-phthalate diesters (58), purportedly safe replacements were developed and introduced into the U.S. market before or in the early 2000s, including di(isononyl) cyclohexane-1,2-dicarboxylate (DiNCH) (59) and a terephthalate diester, di(2-ethylhexyl) terephthalate (DEHTP) (60, 61). Recent observational evidence indicates that DiNCH and DEHTP exposure may be associated with adverse health outcomes, including increased risk of uterine fibroids and pre-term birth (62, 83), as well as altered sex steroid hormone and oxidative stress levels (84-87). Therefore, additional studies are needed to identify maternal characteristics associated with phthalate and replacement exposures, which can be used as covariates in studies evaluating the health implications of increasing and decreasing exposure to these chemicals during pregnancy. To identify sub-populations of pregnant women with higher phthalate/replacement exposures, numerous prior studies evaluated important seasonal, sociodemographic, and lifestyle predictors of phthalate metabolite (and to a lesser extent phthalate replacement) concentrations. In general, these studies evaluated bivariable and/or 11 multivariable associations between predictors and concentrations of individual phthalate biomarkers (as individual biomarkers or molar sums of biomarkers from common parents) and found that the following characteristics most often remained as important determinants of phthalate biomarker concentrations: age, race/ethnicity, education, income, marital status, social class, parity, pre-pregnancy body mass index (BMI), smoking status, diet, and study year (17, 18, 88-93). However, these studies only assessed determinants of individual biomarkers (or molar sums representing single parent compounds) in single-pollutant models. Given that pregnant women are exposed to numerous phthalates and their replacements, some studies also assessed predictors of maternal exposure to numerous chemicals (including phthalates) using unsupervised learning methods such as k-means clustering and principal component analysis (PCA) (94-97). One study using k-means paired with logistic regression analyses found that race/ethnicity and diet were associated with clusters of women who had higher biomarker concentrations of personal care product and plasticizer phthalates, respectively (95). Another study observed associations of parity, pre-pregnancy BMI, and job type with high phthalate cluster membership (97). The few studies using PCA paired with linear regression analyses reported that age, birthplace, race/ethnicity, income, job type, parity, pre-pregnancy BMI, smoking status, and diet were associated with principal components (PCs) heavily loaded for phthalate metabolites, including mono(3-carboxypropyl) phthalate (MCPP), mono-n-butyl phthalate (MBP), monobenzyl phthalate (MBzP), mono- iso-butyl phthalate (MiBP), and monoethyl phthalate (MEP), and metabolites of di(2- ethylhexyl) phthalate (DEHP) (94, 95, 97). Such approaches that identify determinants of biomarker concentration patterns/profiles may better identify unique characteristics of 12 women who may benefit most from interventions targeted at decreasing phthalate/replacement exposure. Given that phthalates and their replacements are a diverse class of chemicals with multiple exposure sources, our study focused on identifying whether maternal sociodemographic characteristics, lifestyle factors, enrollment year, and conception season are predictors of phthalate/replacement biomarker concentrations. Our first objective was to ascertain patterns of phthalate/replacement biomarker concentrations using both k-means clustering and PCA, which can identify groups of pregnant women with similar phthalate/replacement biomarker concentration profiles (k-means) and groups of phthalates/replacements that likely exist together (PCA). Our second objective was to evaluate associations of maternal sociodemographic characteristics, early gestation lifestyle factors, conception season, and study enrollment year with the patterns of phthalate/replacement biomarker concentrations identified in k-means and PCA. 2.4. MATERIALS AND METHODS 2.4.1. Illinois Kids Development Study (I-KIDS) recruitment and enrollment The current study includes pregnant women from I-KIDS, an ongoing prospective pregnancy cohort designed to evaluate the impacts of prenatal environmental chemical exposures on infant neurodevelopment. Pregnant women were recruited at their first prenatal care appointment from two local obstetric clinics in Champaign-Urbana, IL. Women who expressed interest in the study were eligible to participate if they were ≥ 10 but < 15 weeks pregnant, 18-40 years old, fluent in English, in a low-risk singleton 13 pregnancy, living within a 30-minute drive of the University of Illinois campus, and not planning to move out of the area before their child’s first birthday. The current study includes the first 482 women who enrolled in I-KIDS between December 2013 and August 2018, and remained in the study through the birth of their infant. These women provided written informed consent and the study was approved by the Institutional Review Board at the University of Illinois. The analysis of de-identified specimens at the Centers for Disease Control and Prevention (CDC) laboratory was determined not to constitute engagement in human subjects research. 2.4.2. Collection of maternal sociodemographic, lifestyle, and conception season information Immediately after enrollment, an I-KIDS staff member visited each participant’s home to obtain information about sociodemographic and lifestyle characteristics. We collected information about the following sociodemographic characteristics using an interviewer- administered questionnaire: age, race/ethnicity, education level, marital status, employment status, and household annual income. Women additionally reported whether they planned their current pregnancy. To determine conception season, we used the estimated due date based on the first day of the last menstrual period reported at baseline and confirmed after the first trimester ultrasound. Each woman reported the following information since conception: smoking status, the number of eight-ounce cups of caffeinated beverages consumed on a typical day, and the number of servings of alcoholic beverages consumed per week. Self-reported pre-pregnancy weight and height were used to calculate pre-pregnancy BMI (in kg/m2). Self-reported pre-pregnancy BMI is highly correlated with first trimester measured BMI in other pregnant populations (98- 14 100), as well as ours (r = 0.99, data not shown). Participants completed a semi- quantitative food frequency questionnaire (FFQ) at enrollment that was adapted for pregnant women from the full-length Block-98 FFQ (NutritionQuest, Berkeley, CA) and asked about maternal diet during the previous three months (101). Reported dietary intakes were used to calculate first trimester Alternative Healthy Eating Index 2010 (AHEI- 2010) – an 11-component diet quality measure (scored out of 110) based on food/nutrients predictive of chronic disease risk and mortality; higher scores reflect better diet quality (102, 103). 2.4.3. Assessment of urinary phthalate/replacement biomarker concentrations I-KIDS participants provided up to five first-morning urine samples at the following gestational timepoints: 8-15, 13-22, 19-28, 25-33, 32-40 weeks gestation (median 13, 17, 23, 28, and 34 weeks gestation, respectively) as described previously (86), which corresponded with study home visits (at median 13, 17, and 34 weeks gestation) or routine prenatal care visits (median 23 and 28 weeks gestation). Most women contributed all five urine samples (94.4%), whereas 5.2% and 0.4% contributed four and three urine samples, respectively. Urine samples were collected in polypropylene urine cups and refrigerated immediately. Within 24 hours of collection, urine samples were aliquoted for long-term storage or pooled from each timepoint. Beginning with the first visit’s sample, we added 900 µL of urine to a 5 mL cryovial tube. Each time women provided a sample, we layered fresh urine onto frozen urine from prior gestational timepoints before immediately freezing it at -80° C. At the end of pregnancy, we thawed and vortex all pooled samples to measure specific gravity. For quality assurance and control, we also 15 collected duplicates and purified water blanks every 10 samples to be analyzed at the CDC. We stored all aliquoted urine at -80° C and sent pooled samples on dry ice to the CDC laboratory in three batches in chronological order of enrollment (batch one enrolled December 2013 - February 2015, batch two enrolled February 2015 - July 2016, and batch three enrolled July 2016 - August 2018). The following phthalate/replacement metabolites were quantified in all batches using previously published methods (104, 105): mono(2-ethylhexyl) phthalate (MEHP), mono(2-ethyl-5-hydroxyhexyl) phthalate (MEHHP), mono(2-ethyl-5-oxohexyl) phthalate (MEOHP), mono(2-ethyl-5- carboxypentyl) phthalate (MECPP), monoisononyl phthalate (MiNP), monocarboxyoctyl phthalate (MCOP), monocarboxynonyl phthalate (MCNP), MCPP, MBzP, MEP, MBP, mono-hydroxybutyl phthalate (MHBP), MiBP, mono-hydroxy-isobutyl phthalate (MHiBP), cyclohexane-1,2-dicarboxylic acid-mono(carboxyoctyl) ester (MCOCH), and cyclohexane-1,2-dicarboxylic acid-monohydroxy isononyl ester (MHiNCH). Three additional metabolites were added to the CDC analytical panel for women in batches two and three (monooxononyl phthalate (MONP), mono(2-ethyl-5-hydroxyhexyl) terephthalate (MEHHTP), and mono(2-ethyl-5-carboxypentyl) terephthalate (MECPTP)) (104, 106). The CDC laboratory has rigorous quality control/quality assurance protocols with excellent long-term reproducibility of most phthalate metabolite biomarkers over 3 and 8 month periods and intra- and inter-day coefficients of variation < 14% for most biomarkers (104-106). 2.4.4. Statistical analysis For phthalate/replacement metabolite concentrations below the limit of detection (LOD), 16 we used instrumental-reading values to avoid bias associated with imputing values below the LOD (107). Across the individual and molar sum biomarkers described below, only one woman had a zero concentration for the sum of di(isononyl) cyclohexane-1,2- dicarboxylate metabolites (∑DiNCH) (meaning that her urinary concentrations of both MCOCH and MHiNCH were zero). In final statistical models we added a constant (1.0) to ∑DiNCH before ln-transformation to avoid undefined estimates (108). To account for urine dilution, we adjusted all urinary phthalate/replacement metabolite concentrations using the following formula: Pc = P[(1.016 − 1)/(SG − 1)], where Pc is the specific gravity- adjusted metabolite concentration, P is the measured metabolite concentration (ng/mL), 1.016 is the population median specific gravity, and SG is the specific gravity of each individual urine sample (109). We summed the molar concentrations (in nmol/mL) of the following metabolites to create biomarkers of exposure to phthalate/replacement parent compounds that are metabolized and excreted as multiple urinary metabolites (86): MEHP, MEHHP, MEHOP, and MECPP for the sum of di(2-ethylhexyl) phthalate metabolites (∑DEHP); MiNP and MCOP for the sum of di-isononyl phthalate metabolites (∑DiNP); MBP and MHBP for the sum of di-n-butyl phthalate metabolites (∑DBP); MiBP and MHiBP for the sum of di-iso-butyl phthalate metabolites (∑DiBP); MHiNCH and MCOCH for ∑DiNCH; and MEHHTP and MECPTP for the sum of di(2-ethylhexyl) terephthalate metabolites (∑DEHTP). We also created another ∑DiNP (∑DiNP2) limited to women enrolled between February 2015 and August 2018 to include MONP. These molar concentrations were converted to ng/mL by multiplying ∑DEHP, ∑DiNP (both versions), ∑DBP, ∑DiBP, ∑DiNCH, and ∑DEHTP by the molecular weights of MECPP, MCOP, MBP, MiBP, MHiNCH, and MECPTP, respectively. We estimated exposure to di- 17 isodecyl phthalate, di-n-octyl phthalate, benzylbutyl phthalate, and diethyl phthalate using ng/mL concentrations of their corresponding urinary metabolites MCNP, MCPP, MBzP, and MEP, respectively. To understand how phthalate/replacement metabolite concentrations in I-KIDS compare to those in the general U.S. population, we used data from the National Health and Nutrition Examination Survey (NHANES) survey cycles 2013-14, 2015-16, and 2017-18 (110-112). These NHANES survey cycles correspond with urine collection years in I- KIDS. Though most women in NHANES were not pregnant, we subset the NHANES sample to only include 18 – 40 year-old females with data on urinary phthalate/replacement metabolite concentrations. Finally, because NHANES does not provide specific gravity information, we reported median (25th, 75th percentiles) unadjusted phthalate/replacement metabolite concentrations for both samples (Table 2). Our analyses included 15 maternal characteristics that have been previously shown to predict phthalate/replacement biomarker concentrations or were hypothesized to be critical determinants of phthalate/replacement exposure in our population (17, 88, 91, 113-115). These included age, race/ethnicity, education, marital status, employment status, household annual income, parity, conception season, enrollment year, smoking in the first trimester, consumption of alcohol and caffeine in the first trimester, pregnancy intention, pre-pregnancy BMI, and diet quality. Almost all predictors were assessed as categorical variables, with the exception of enrollment year, which we evaluated as a continuous variable that can be interpreted for every 1 year increase. Details about 18 variable operationalization are provided in Table 1. Of note, an additional category for smoking in the first trimester (“unknown”) was created to account for missingness due to an ambiguous skip pattern in the first iteration of the survey. Pre-pregnancy BMI was categorized based on standard U.S. clinical cut-offs (116). We selected methods to evaluate chemical mixtures appropriate for the specific research question (66). We used the following two unsupervised methods (objective 1): k-means clustering and PCA (68, 69). K-means clustering identifies subgroups of women with distinct biomarker concentration profiles, which is useful for identifying pregnant women who may experience relatively high or low chemical exposures. We used k-means clustering to group pregnant women into k number of distinct, non-overlapping clusters (identified using Euclidean geometry) based on their similarities across all individual phthalate/replacement biomarker concentrations. To identify the optimal number of clusters, we compared 1, 2, 3, and 4 cluster solutions using the pseudo f-statistic index (the ratio of between-cluster variance to within cluster variance) and confirmed the ideal number of clusters using elbow plots of R2 values. PCA identifies linear combinations of biomarker concentration patterns among highly correlated biomarker that explain most of the variance in biomarker concentrations in a population. These resulting patterns can be related to common exposure sources or behaviors in the study population. We used PCA with a Varimax rotation to identify biomarkers of highly correlated phthalate/replacements to which pregnant women are likely exposed and created distinct, uncorrelated PC scores that explain most of the variance in phthalate/replacement biomarker concentrations in our participants. To determine the ideal number of PCs, we assessed elbow plots of 19 eigenvalues (total variance explained by each component) and used the total variance explained to confirm the optimal number of components that best represents the data. We considered biomarkers with loadings ≥ 0.3 to be notable. For both k-means and PCA, we included specific gravity-adjusted phthalate/replacement biomarker concentrations in ng/mL that were ln-transformed and z-transformed. We used logistic and linear regression models to evaluate associations of 15 maternal characteristics with the identified clusters and PCs, respectively (objective 2). Evaluating associations of characteristics with identified k-means clusters using logistic regression models provides information about characteristics of pregnant women with specific phthalate/replacement biomarker concentration profiles. Assessing relationships of maternal characteristics with identified PCs using linear regression models provides information about characteristics that likely result in exposure to certain phthalates/replacements from common exposure sources or behaviors. We evaluated both unadjusted models (bivariable analyses) and models simultaneously adjusted for all 15 predictors (multivariable analyses). A total of 9 women had missing data on at least one predictor. Therefore, 473 women who enrolled between December 2013 and August 2018 (referred to as the full sample) were included in final multivariable analyses. To assess MONP and both DEHTP metabolites, we conducted additional analyses limited to women enrolled between February 2015 and August 2018 (referred to as the sub- sample). A total of 305 women were included in these multivariable analyses. There was high agreement between the full and sub-samples with regards to k-means cluster membership (Kappa statistic = 0.82) and PC scores (r > 0.8). However, we reported 20 results from both samples to provide information about phthalate/∑DiNCH biomarker concentrations across the whole study period and to report results related to phthalate replacement DEHTP. We used SAS 9.4 (version 15.1, SAS Institute) for all statistical analyses. We used PROC FASTCLUS and PROC LOGISTIC to assign k-means clusters and for bivariable and multivariable logistic regression models, respectively. We used PROC FACTOR for the PCA and PROC GENMOD for bivariable and multivariable linear regression models. Based on recommendations from the American Statistical Association and others (117, 118), rather than using P-values, we used the magnitude of associations and 95% confidence intervals (CIs) to identify potentially meaningful results. We used RStudio Version 1.3.1093 (RStudio, Boston, MA) to generate figures. 2.5. RESULTS 2.5.1. I-KIDS characteristics and urinary phthalate/replacement biomarker concentrations Sociodemographic and lifestyle characteristics of I-KIDS women have been previously described (86) and are outlined in Table 1. Briefly, most women were non-Hispanic white, of high socioeconomic status, and engaged in healthy lifestyle behaviors. Greater than 97% of I-KIDS women had detectable urinary concentrations of at least one metabolite per phthalate parent compounds (including DEHTP), while only 77% had detectable urinary concentrations of at least one DiNCH metabolite (Table 2). Most phthalate/replacement biomarkers were weakly-to-moderately correlated (r < 0.4), 21 although strong correlations were observed between MCPP, ∑DiNP, and ∑DiNP2 (r > 0.8; Figure 2). I-KIDS pregnant women had similar median urinary phthalate and DiNCH metabolite concentrations as those from a nationally representative sample of 18 – 40 year-old pregnant or non-pregnant U.S. women from the 2013 - 2018 National Health and Nutrition Examination Survey (NHANES) cycles (Table 2). However, I-KIDS women had higher median concentrations of DEHTP metabolites, but lower median concentrations of MEP (with overlapping 25th and 75th percentiles) compared to NHANES women. 22 Table 1. Characteristics of I-KIDS women in the full and sub-samples. Full sample enrolled Sub-sample enrolled 12/2013 - 8/2018 2/2015 - 8/2018 (n=482) (n=309) n (%) n (%) Maternal age < 30 years 197 (41.9) 115 (37.2) ≥ 30 years 285 (59.1) 194 (62.8) Race/ethnicity 1 missing Non-Hispanic white 385 (80.0) 251 (81.2) Non-Hispanic black 26 (5.4) 16 (5.2) Asian 26 (5.4) 20 (6.5) Other1 44 (9.2) 22 (7.1) Education Some college or less 90 (18.7) 52 (16.8) College grad or higher 392 (81.3) 257 (83.2) Marital status Married 426 (88.4) 273 (88.4) Unmarried 56 (11.6) 36 (11.7) Employment status Unemployed 67 (13.9) 39 (12.6) Employed 415 (86.1) 270 (87.4) Household income 4 missing <$60,000 138 (28.6) 84 (27.2) $60,000-$99,999 182 (37.8) 114 (36.9) ≥$100,000 158 (32.8) 109 (35.3) Enrollment year 12/2013 – 02/2015 173 (35.9) -- 02/2015 – 07/2016 174 (36.1) 174 (56.3) 07/2016 – 08/2018 135 (28.0) 135 (43.7) Conception season Winter 122 (25.3) 80 (25.9) Spring 134 (27.8) 101 (32.7) Summer 107 (22.2) 78 (25.2) Fall 119 (24.7) 50 (16.2) Parity 0 children 246 (51.0) 164 (53.1) ≥ 1 child 236 (49.0) 145 (46.9) Active smoker No 423 (87.8) 294 (95.2) Yes 24 (5.0) 15 (4.9) Missing 35 (7.3) -- Alcohol intake 1 missing No serving/week 281 (58.3) 180 (58.3) ≥ 1 servings/week 200 (41.5) 129 (41.8) Pregnancy intention Planned 321 (66.6) 207 (67.0) Unplanned 161 (33.4) 102 (33.0) Pre-pregnancy BMI < 25 kg/m2 258 (53.5) 162 (52.4) ≥ 25 kg/m2 224 (46.5) 147 (47.6) Caffeine intake None 196 (40.7) 124 (40.1) < 1 cups/week 145 (30.1) 93 (30.1) ≥ 1 cups/week 141 (29.3) 92 (29.7) Overall diet quality 3 missing 2 missing AHEI < 55.1 239 (49.6) 153 (49.5) AHEI ≥ 55.1 240 (49.8) 154 (49.8) 1 Hispanic white, American Indian or Alaska Native, Native Hawaiian or other Pacific Islander, multiracial, and other. I-KIDS, Illinois Kids Development Study. 23 Table 2. Unadjusted phthalate/replacement metabolite concentrations (ng/mL). I-KIDS (n=482) NHANES (n=1076) Parent Median (25th, 75th pctl) Median (25th, 75th pctl) Metabolite(s) % ≥ LOD Compound 2013-2018 2013-2018 MEHP 74.3 1.3 (0.8, 2.2) 1.1 (0.6, 2.3) MEHHP 100.0 6.0 (3.8, 9.2) 5.3 (2.5, 10.8) DEHP MEOHP 100.0 4.6 (3, 6.9) 3.7 (1.7, 7.3) MECPP 100.0 9.2 (6.1, 14.8) 8.6 (4.1, 16.8) MCOP 100.0 11.0 (5.4, 25.7) 8.0 (3.5, 23.6) DiNP MiNP 41.7 0.7 (0.4, 1.5) 0.6 (0.6, 1.1) MONP 100.0 2.7 (1.7, 4.7)1 1.6 (0.7, 3.2)3 DiDP MCNP 100.0 2.1 (1.4, 3.3) 1.6 (0.8, 3.4) DOP MCPP 97.1 1.5 (0.9, 2.6) 1.1 (0.5, 2.6) BBzP MBzP 99.6 5.3 (2.8, 12) 4.6 (1.6, 12.0) DEP MEP 100.0 25.0 (12.6, 46.5) 34.4 (14.6, 85.8) MBP 100.0 12.6 (8.1, 19.5) 11.2 (5.1, 20.2) DBP MHBP 90.0 1.2 (0.7, 2) 0.9 (0.3, 1.7)2 MiBP 99.8 9.1 (5.5, 14.1) 8.7 (4.0, 17.9) DiBP MHiBP 99.8 3.3 (2, 5.1) 2.9 (1.4, 6.0)2 MHiNCH 77.4 0.8 (0.4, 1.6) 0.4 (0.3, 1.1) DiNCH MCOCH 50.4 0.5 (0.3, 1) 0.4 (0.4, 0.8)3 MEHHTP 100.0 8.7 (3.7, 19.7)1 6.0 (2.2, 17.1)3 DEHTP MECPTP 100.0 60.5 (24.3, 140)1 20.7 (8.5, 67.6)3 Urinary phthalate/replacement metabolite concentrations were obtained for 18-40 year-old pregnant and non-pregnant females from NHANES survey years 2013-14, 2015-16, and 2017-18 I-KIDS reports numeric values for all concentrations below the LOD, while NHANES replaces all values below the LOD with the LOD/√2 for that metabolite. Concentrations do not account for urine dilution. 1n=309, 2n=1074, 3n=682. I-KIDS, Illinois Kids Development Study; NHANES, National Health and Nutrition Examination Survey. 24 Figure 2. Correlations between urinary specific gravity-adjusted phthalate/replacement biomarker concentrations. Heat map presents Pearson correlations between all biomarkers. Yellow and turquoise indicate negative and positive correlations, respectively, where lighter shades represent weaker correlations and darker shades represent stronger correlations. 25 2.5.2. K-means clusters of women with distinct phthalate/replacement biomarker concentration profiles Our goal with using k-means clustering was to identify groups of women with distinct profiles of urinary phthalate/replacement biomarker concentrations. In the full sample (excluded MONP and the DEHTP metabolites), we identified the following two clusters: cluster 1 included women with concentrations of all phthalate biomarkers below the sample median, while cluster 2 included women with concentrations of all phthalate biomarkers above the sample median (Figure 3). ∑DiNCH concentrations were similar between the two clusters, and therefore did not drive cluster membership. In the sub- sample (includes MONP and DEHTP metabolites), the k-means procedure identified similar clusters as those identified in the full sample. Therefore, the two clusters of women in this sub-sample included women with all phthalate biomarker concentrations (including ∑DEHTP) below the sample median (cluster 1) and those with all phthalate biomarker concentrations (including ∑DEHTP) above the sample median (cluster 2) (Figure 3). 26 Figure 3. Multivariable associations of sociodemographic, lifestyle, enrollment year, and conception season predictors with k-means clusters. Median urinary specific gravity-adjusted phthalate/replacement biomarker concentrations by k-means cluster in the A) full sample (enrolled between 12/2013 and 8/2018) or C) sub-sample (enrolled between 2/2015 and 8/2018). Logistic regression models simultaneously adjusted for all listed predictors evaluated associations of 15 predictors with the odds of having B) high phthalate (cluster 2, n=230) versus low phthalate (cluster 1, n=243) in the full sample or D) high phthalate including ∑DEHTP (cluster 2, n=131) versus low phthalate including ∑DEHTP (cluster 1, n=174) biomarker concentrations in the sub-sample. BMI, body mass index; CI, confidence interval. 27 2.5.3. Associations of maternal characteristics with identified k-means clusters Bivariable analyses evaluating associations of characteristics with k-means clusters for the full and sub-samples are presented in Table 3. In multivariable logistic regression models simultaneously adjusted for all characteristics, women had higher odds of high phthalate cluster membership if they had ≥ 1 child prior to the I-KIDS pregnancy (ref = no children; OR: 1.6; 95% CI: 1.0, 2.6), had overweight or obesity before pregnancy (ref = under-/normal weight; OR: 1.4; 95% CI: 0.9, 2.2), and consumed < 1 cup of caffeine per week (ref = no caffeine consumption; OR: 1.9; 95% CI: 1.1, 3.2) (Figure 3 and Table 3). Conversely, women had lower odds of high phthalate cluster membership if they were ≥ 30 years old (ref = < 30 years; OR: 0.6; 95% CI: 0.4, 1.0), enrolled earlier in the study (for every 1 year increase in study year; OR: 0.5; 95% CI: 0.4, 0.7), and conceived in the spring (OR: 0.5; 95% CI: 0.3, 0.9), summer (OR: 0.3; 95% CI: 0.2, 0.6), or fall (OR: 0.6; 95% CI: 0.4, 1.1) compared to winter. In multivariable logistic regression models in the sub-sample, associations of enrollment year, conception season, and parity with cluster membership were similar to those observed in the full sample (Figure 3 and Table 3). However, additional associations of marital status and annual household income with cluster membership emerged in the multivariable logistic regression models in the sub- sample (Figure 3 and Table 3). Specifically, women had a higher odds of high phthalate (including ∑DEHTP) cluster membership if they were unmarried (ref = married, OR: 2.0; 95% CI: 0.8, 5.1) and had annual household incomes < $100,000 (ref = ≥ $100,000; < $60,000 OR: 1.7; 95% CI: 0.8, 3.5; $60,000 - $99,999 OR: 1.6; 95% CI: 0.9, 2.9). 28 Table 3. Associations of maternal sociodemographic and lifestyle characteristics with k-means clusters. Unadjusted OR (95% CI) Unadjusted OR (95% CI) n=482 n=309 Maternal age < 30 years ref ref ≥ 30 years 0.6 (0.4, 0.9) 0.7 (0.5, 1.2) Race/ethnicity Non-Hispanic White ref ref Non-Hispanic Black 1.7 (0.8, 3.9) 1.8 (0.7, 5.0) Asian 0.9 (0.4, 2.1) 0.9 (0.4, 2.4) Other 1.2 (0.6, 2.2) 2.0 (0.8, 5.0) Education Some college or less 1.6 (1.0, 2.6) 1.1 (0.6, 2.1) College grad or higher ref ref Marital status Married ref ref Unmarried 1.7 (1.0, 3.0) 2.3 (1.1, 4.6) Employment status Unemployed 1.2 (0.7, 1.9) 0.8 (0.4, 1.6) Employed ref ref Household income <$60,000 1.9 (1.2, 3.0) 2.0 (1.1, 3.7) $60,000-$99,999 1.6 (1.0, 2.5) 1.7 (1.0, 2.9) ≥$100,000 ref ref Enrollment year Continuous (1 year increase) 0.5 (0.5, 0.6) 0.8 (0.7, 1.1) Conception season Winter ref ref Spring 0.4 (0.3, 0.7) 0.5 (0.3, 0.9) Summer 0.3 (0.2, 0.6) 0.3 (0.2, 0.6) Fall 0.7 (0.4, 1.1) 0.7 (0.3, 1.4) Parity 0 children ref ref ≥ 1 child 1.7 (1.2, 2.4) 1.5 (0.9, 2.3) Active smoker No ref ref Yes 1.4 (0.6, 3.2) 2.0 (0.7, 5.8) Skip Pattern Missing 7.1 (2.7, 18.6) -- Alcohol intake No serving/week ref ref ≥ 1 servings/week 0.8 (0.6, 1.2) 1.1 (0.7, 1.7) Pregnancy intention Planned 0.6 (0.4, 0.9) 0.6 (0.4, 1.0) Unplanned ref ref Pre-pregnancy BMI < 25 kg/m2 ref ref ≥ 25 kg/m2 1.2 (0.8, 1.7) 1.1 (0.7, 1.8) Caffeine intake None ref ref < 1 cups/week 1.6 (1.0, 2.4) 1.3 (0.8, 2.3) ≥ 1 cups/week 1.1 (0.7, 1.8) 1.3 (0.7, 2.2) 29 Table 3 (cont’d). Unadjusted OR (95% CI) Unadjusted OR (95% CI) n=482 n=309 Overall diet quality AHEI < 55.1 1.3 (0.9, 1.8) 1.3 (0.8, 2.1) AHEI ≥ 55.1 ref ref Data are presented as odds ratios (95% CI) from unadjusted and adjusted logistic regression models evaluating the probability of being in cluster 2 (high phthalate/∑DEHTP) compared to cluster 1 (low phthalate/∑DEHTP). Adjusted models control for all the maternal characteristics listed in the table. CI, confidence interval. Bold indicate potentially meaningful findings. Full sample n=243 and n=230 for clusters 1 and 2, respectively; sub-sample n=174 and n=131 for clusters 1 and 2, respectively. 2.5.4. PCs of phthalate/replacement biomarker concentrations Our goal with PCA was to identify phthalate/replacement biomarker concentrations that exist together due to common exposure sources or lifestyle factors. In the full sample, four PCs accounted for 71.2% of the total variance (32.2%, 17.2%, 11.2%, and 10.6% of the total variance explained by components 1-4, respectively). The heaviest loadings for each PC were as follows: ∑DiNP, MCNP, and MCPP with component 1 (referred to as phthalate plasticizer component); MBzP, ∑DBP, and ∑DiBP with component 2 (referred to as other phthalate component); ∑DEHP and ∑DiNCH with component 3 (referred to as ∑DEHP/∑DiNCH component); and MEP with component 4 (referred to as MEP component) (Table 4). In the sub-sample, four PCs accounted for 66.9% of the total variance (27.6%, 16.3%, 13.1%, and 9.9% of the total variance explained by components 1-4, respectively). Components 2 (other phthalate component) and 4 (MEP component) were similar to those discussed above, whereas component 1 was heavily loaded by ∑DiNP2, MCNP, and MCPP (referred to as plasticizer phthalate component) and component 3 was heavily loaded by ∑DiNCH and ∑DEHTP (referred to as phthalate replacement component) (Table 4). In the full sample and the sub-sample, the biomarkers that loaded most heavily were positively correlated with the four component scores indicating that as urinary concentrations of those biomarkers increase, component scores 30 increase (Table 4). Table 4. Standardized scoring coefficients from principal component analyses. Biomarker Component 1 Component 2 Component 3 Component 4 Full sample (enrolled 12/2013 - 08/2018, n=482) ∑DEHP 0.18615 0.12433 0.30419 -0.01715 ∑DiNP 0.38358 -0.11581 -0.04215 -0.00094 MCNP 0.32727 -0.07695 -0.08505 -0.01991 MCPP 0.35948 -0.02074 0.0854 -0.00803 MBzP -0.05799 0.39207 -0.20075 0.02746 MEP -0.0205 -0.05858 -0.00947 0.99756 ∑DBP -0.0426 0.45189 0.03671 0.01679 ∑DiBP -0.09515 0.44958 0.04137 -0.14473 ∑DiNCH 0.02608 -0.04767 0.90648 -0.0018 Sub-sample (enrolled 02/2015 - 08/2018, n=309) ∑DEHP 0.09557 0.17086 0.02263 0.15387 ∑DiNP2 0.43855 -0.13006 -0.04165 -0.00371 MCNP 0.33061 -0.01817 -0.04642 -0.02653 MCPP 0.39854 -0.00695 -0.03763 -0.08151 MBzP -0.05536 0.34434 0.01587 0.14104 MEP -0.04751 -0.04257 -0.02487 0.91181 ∑DBP -0.05685 0.44587 0.01462 -0.00131 ∑DiBP -0.06325 0.44524 -0.04736 -0.30297 ∑DiNCH -0.09997 0.03464 0.6192 -0.09097 ∑DEHTP 0.001 -0.03457 0.56354 0.06057 Bold indicate biomarkers with notable loadings (≥ 0.3). 2.5.5. Associations of maternal characteristics with identified PCs Results of bivariable analyses evaluating the relationships between characteristics and PC scores for the full and sub-samples are presented in Tables 5 and 6, respectively. In multivariable linear regression models in the full sample (Figure 4 and Table 5), phthalate plasticizer component scores were lower in Asian women (ref = non-Hispanic white), those with annual household incomes < $60,000 (ref = ≥ $100,000), those who conceived in the spring, summer, or fall (ref = winter), and those who planned their pregnancy (ref = unplanned pregnancy). Conversely, plasticizer component scores were higher in women who had lower educational attainment (ref = college graduates), enrolled earlier in the 31 study, those who had overweight or obesity before pregnancy (ref = under-/normal weight), and who consumed < 1 cups of caffeine/week (ref = no caffeine consumption). Other phthalate scores were lower in women ≥ 30 years old (ref = < 30 years old) and those with lower educational attainment (ref = college graduates), but higher in Asian women or women of other race/ethnicity (ref = non-Hispanic white), those with annual household incomes < $100,000 (ref = ≥ $100,000), and those who had ≥ 1 child prior to the I-KIDS pregnancy (ref = no children). ∑DEHP/∑DiNCH component scores were lower in women who conceived in the spring or summer months (ref = winter), in those with annual household incomes < $60,000 (ref = ≥ $100,000), and in women with unknown smoking status (ref = non-smokers). However, ∑DEHP/∑DiNCH component scores were higher in Asian women or those of other race/ethnicity (ref = non-Hispanic white), in women who enrolled later in the study, in those who smoked in the first trimester (ref = non-smokers), and in those that had overweight or obesity before pregnancy (ref = under- /normal weight). Lastly, MEP scores were higher in black women or those of other race/ethnicity (ref = non-Hispanic white), in unmarried women (ref = married), and among those who consumed some amount of caffeine/week (ref = no caffeine consumption), while MEP scores were lower in women who conceived in the spring or fall months (ref = winter), who consumed ≥ 1 servings/week of alcohol in the first trimester (ref = no alcohol consumption), who had overweight or obesity before pregnancy (ref = under-/normal weight), and had poor first trimester diet quality (ref = better diet quality). 32 Table 5. Associations of predictors with 4 principal component scores in the full sample. Component 1 Component 2 Component 3 Component 4 Unadjusted Unadjusted Unadjusted Unadjusted β (95% CI) β (95% CI) β (95% CI) β (95% CI) Maternal age < 30 years ref ref ref ref ≥ 30 years -0.1 (-0.3, 0.1) -0.2 (-0.4, 0.0) 0.2 (0.0, 0.4) 0.0 (-0.2, 0.2) Race/ethnicity Non-Hispanic White ref ref ref ref Non-Hispanic Black 0.1 (-0.3, 0.5) 0.1 (-0.3, 0.5) -0.1 (-0.5, 0.3) 1.3 (0.9, 1.7) Asian -0.5 (-0.9, -0.1) 0.5 (0.1, 0.9) 0.2 (-0.2, 0.6) 0.1 (-0.3, 0.4) Other 0.1 (-0.2, 0.4) 0.2 (-0.1, 0.5) 0.4 (0.1, 0.7) 0.5 (0.2, 0.8) Education Some college or less 0.2 (0.0, 0.5) 0.1 (-0.1, 0.3) -0.1 (-0.3, 0.1) 0.2 (0.0, 0.4) College grad or ref ref ref ref higher Marital status Married ref ref ref ref Unmarried 0.1 (-0.2, 0.4) 0.2 (-0.1, 0.5) 0.0 (-0.3, 0.3) 0.6 (0.3, 0.9) Employment status Unemployed 0.0 (-0.3, 0.2) 0.3 (0.0, 0.5) 0.0 (-0.3, 0.3) 0.1 (-0.1, 0.4) Employed ref ref ref ref Household income < $60,000 0.0 (-0.3, 0.2) 0.6 (0.3, 0.8) -0.3 (-0.5, -0.1) 0.2 (0.0, 0.4) $60,000-$99,999 0.0 (-0.2, 0.2) 0.2 (0.0, 0.4) -0.2 (-0.4, 0.0) 0.1 (-0.1, 0.3) ≥ $100,000 ref ref ref ref Enrollment year Continuous -0.3 (-0.4, -0.3) 0.0 (-0.1, 0.1) 0.3 (0.2, 0.4) 0.0 (-0.1, 0.1) Conception season Winter ref ref ref ref Spring -0.6 (-0.8, -0.3) 0.1 (-0.1, 0.4) 0.0 (-0.3, 0.2) -0.2 (-0.4, 0.1) Summer -0.6 (-0.9, -0.4) 0.0 (-0.2, 0.3) -0.1 (-0.4, 0.1) 0.0 (-0.2, 0.3) Fall -0.3 (-0.5, -0.1) 0.1 (-0.1, 0.4) 0.0 (-0.3, 0.2) -0.2 (-0.4, 0.1) Parity 0 children ref ref ref ref ≥ 1 child 0.2 (0.0, 0.3) 0.3 (0.1, 0.5) 0.1 (-0.1, 0.3) 0.1 (-0.1, 0.2) Active smoker No ref ref ref ref Yes 0.4 (-0.1, 0.8) 0.1 (-0.3, 0.5) 0.2 (-0.2, 0.6) 0.4 (-0.1, 0.8) Skip Pattern Missing 0.7 (0.3, 1.0) 0.1 (-0.2, 0.5) -0.8 (-1.1, -0.4) -0.1 (-0.4, 0.3) Alcohol intake No serving/week ref ref ref ref ≥ 1 servings/week 0.0 (-0.2, 0.2) -0.1 (-0.3, 0.1) 0.1 (-0.1, 0.3) -0.1 (-0.3, 0.1) Pregnancy intention Planned -0.2 (-0.4, 0.0) -0.2 (-0.4, 0.0) -0.1 (-0.3, 0.1) -0.1 (-0.3, 0.1) Unplanned ref ref ref ref Pre-pregnancy BMI < 25 kg/m2 ref ref ref ref ≥ 25 kg/m2 0.1 (-0.1, 0.2) 0.1 (-0.1, 0.2) 0.1 (-0.1, 0.3) -0.1 (-0.3, 0.1) 33 Table 5 (cont’d). Component 1 Component 2 Component 3 Component 4 Unadjusted Unadjusted Unadjusted Unadjusted β (95% CI) β (95% CI) β (95% CI) β (95% CI) Caffeine intake None ref ref ref ref < 1 cups/week 0.2 (0.0, 0.4) 0.0 (-0.2, 0.2) 0.1 (-0.1, 0.3) 0.2 (0.0, 0.4) ≥ 1 cups/week 0.1 (-0.1, 0.3) 0.0 (-0.3, 0.2) 0.2 (-0.1, 0.4) 0.2 (0.0, 0.4) Overall diet quality AHEI < 55.1 0.1 (-0.1, 0.3) 0.0 (-0.2, 0.2) 0.0 (-0.2, 0.1) 0.0 (-0.2, 0.2) AHEI ≥ 55.1 ref ref ref ref Data are presented as β-estimates (95% CI) from unadjusted (n=482) and adjusted (n=473) linear regression models. Adjusted models control for all the maternal characteristics listed in the table. CI, confidence interval. Bold indicate potentially meaningful findings. In multivariable analyses in the sub-sample (Figure 4 and Table 6), associations of race/ethnicity, enrollment year, and conception season with the phthalate plasticizer component, of race/ethnicity, annual household income, and parity with the other phthalate component, of annual household income, enrollment year, conception season, and pre-pregnancy BMI with the phthalate replacement component, and of race/ethnicity, marital status, and conception season with the MEP component were similar to those reported above in the full sample. However, unique associations of maternal characteristics with each component also emerged in this sub-sample. Phthalate plasticizer scores were higher among women who were unemployed (ref = employed) and those who smoked in the first trimester (ref = non-smoker). Other phthalate scores were higher among women who enrolled later in the study, but were lower in those who had lower educational attainment (ref = college graduates) and had poor first trimester diet quality (ref = better diet quality). Phthalate replacement scores were lower among women who were unemployed (ref = employed), but higher among women ≥ 30 years old (ref = < 30 years old). Lastly, MEP scores were higher among women ≥ 30 years old (ref = < 30 years old), those who enrolled earlier in the study, those who smoked in the first 34 trimester (ref = non-smoker), and those who planned their current pregnancy (ref = unplanned pregnancy). Figure 4. Multivariable associations of sociodemographic, lifestyle, enrollment year, and conception season predictors with principal component scores. Linear regression models simultaneously adjusted for all listed predictors evaluated associations of 15 maternal predictors with change in component scores in the A) full sample (enrolled between 12/2013 and 8/2018, n=473) and B) sub-sample (enrolled between 2/2015 and 8/2018, n=305). Component 1 = phthalate plasticizer component, component 2 = other phthalate component, component 3 = ∑DEHP/∑DEHTP (full sample) or phthalate replacement (sub-sample) component, and component 4 = MEP component. BMI, body mass index; CI, confidence interval. 35 Table 6. Associations of predictors with 4 principal component scores in the sub- sample. Component 1 Component 2 Component 3 Component 4 Unadjusted Unadjusted Unadjusted Unadjusted β (95% CI) β (95% CI) β (95% CI) β (95% CI) Maternal age < 30 years ref ref ref ref ≥ 30 years 0.0 (-0.3, 0.2) 0.0 (-0.3, 0.2) 0.1 (-0.1, 0.4) 0.1 (-0.1, 0.4) Race/ethnicity Non-Hispanic White ref ref ref ref Non-Hispanic Black 0.1 (-0.4, 0.6) -0.1 (-0.5, 0.5) 0.2 (-0.3, 0.7) 1.2 (0.8, 1.7) Asian -0.6 (-1.0, -0.1) 0.7 (0.3, 1.1) -0.3 (-0.7, 0.2) 0.1 (-0.4, 0.5) Other -0.2 (-0.6, 0.2) 0.4 (0.0, 0.8) 0.3 (-0.1, 0.8) 0.7 (0.3, 1.1) Education Some college or less 0.0 (-0.3, 0.3) -0.1 (-0.4, 0.2) 0.2 (-0.1, 0.5) 0.3 (0.01, 0.6) College grad or higher ref ref ref ref Marital status Married ref ref ref ref Unmarried 0.0 (-0.4, 0.3) 0.1 (-0.2, 0.5) 0.2 (-0.1, 0.6) 0.7 (0.4, 1.1) Employment status Unemployed -0.3 (-0.6, 0.1) 0.3 (-0.1, 0.6) -0.3 (-0.6, 0.0) 0.2 (-0.2, 0.5) Employed ref ref ref ref Household income <$60,000 -0.1 (-0.3, 0.2) 0.6 (0.3, 0.8) -0.1 (-0.4, 0.2) 0.2 (-0.1, 0.5) $60,000-$99,999 -0.1 (-0.4, 0.2) 0.3 (0.04, 0.6) -0.1 (-0.3, 0.2) 0.1 (-0.2, 0.3) ≥$100,000 ref ref ref ref Enrollment year Continuous -0.1 (-0.2, 0.0) 0.0 (-0.1, 0.2) 0.2 (0.1, 0.3) 0.0 (-0.2, 0.1) Conception season Winter ref ref ref ref Spring -0.5 (-0.8, -0.3) 0.1 (-0.2, 0.4) -0.1 (-0.4, 0.2) -0.3 (-0.6, 0.0) Summer -0.6 (-0.9, -0.3) -0.1 (-0.4, 0.2) -0.3 (-0.6, 0.0) -0.1 (-0.4, 0.2) Fall -0.4 (-0.7, -0.1) 0.1 (-0.2, 0.5) -0.1 (-0.5, 0.3) -0.2 (-0.5, 0.2) Parity 0 children ref ref ref ref ≥ 1 child 0.1 (-0.1, 0.3) 0.3 (0.1, 0.5) 0.0 (-0.2, 0.3) 0.1 (-0.1, 0.3) Active smoker No ref ref ref ref Yes 0.4 (-0.1, 0.9) 0.0 (-0.5, 0.5) 0.2 (-0.3, 0.7) 0.6 (0.1, 1.1) Alcohol intake 0 drinks ref ref ref ref > 1 drink 0.0 (-0.2, 0.3) -0.1 (-0.3, 0.2) 0.2 (0.01, 0.5) -0.1 (-0.3, 0.2) Pregnancy intention Planned -0.1 (-0.4, 0.1) -0.2 (-0.4, 0.1) -0.2 (-0.4, 0.1) -0.1 (-0.4, 0.1) Unplanned ref ref ref ref Pre-pregnancy BMI < 25 kg/m2 ref ref ref ref ≥ 25 kg/m2 0.0 (-0.2, 0.2) 0.0 (-0.2, 0.3) 0.3 (0.1, 0.5) -0.1 (-0.3, 0.2) 36 Table 6 (cont’d). Component 1 Component 2 Component 3 Component 4 Unadjusted Unadjusted Unadjusted Unadjusted β (95% CI) β (95% CI) β (95% CI) β (95% CI) Caffeine intake None ref ref ref ref <1 cups/week 0.2 (-0.1, 0.5) 0.1 (-0.2, 0.3) 0.1 (-0.2, 0.3) 0.1 (-0.2, 0.4) ≥1 cups/week 0.0 (-0.3, 0.3) 0.1 (-0.2, 0.4) 0.1 (-0.2, 0.3) 0.2 (-0.1, 0.5) Overall diet quality AHEI < 55.1 0.1 (-0.1, 0.3) -0.2 (-0.4, 0.1) 0.2 (-0.1, 0.4) 0.1 (-0.2, 0.3) AHEI ≥ 55.1 ref ref ref ref Data are presented as β-estimates (95% CI) from unadjusted (n=309) and adjusted (n=305) linear regression models. Adjusted models control for all the maternal characteristics listed in the table. CI, confidence interval. Bold indicate potentially meaningful findings. 2.6. DISCUSSION In the current study, we identified two distinct clusters of women: those with low phthalate (including ∑DEHTP) and those with high phthalate (including ∑DEHTP) biomarker concentrations. We also identified four components representative of phthalate/replacement biomarker concentrations from common exposure sources or those that track with certain behaviors. We identified age, marital status, annual household income, parity, pre-pregnancy BMI, caffeine intake, conception season, and enrollment year as important predictors of k-means clusters and at least one PC. Additionally, race/ethnicity, education, employment, pregnancy intention, smoking and consuming alcohol in the first trimester, and first trimester diet quality were identified as important determinants of at least one PC. Overall, our findings contribute information about predictors of phthalate/replacement mixtures that may be important confounding factors in studies evaluating associations of chemical mixtures with pregnancy-related health outcomes. Furthermore, our results may inform future perinatal health recommendations by providing insights into characteristics of pregnant women who are most likely to be exposed to phthalates and their replacements. 37 2.6.1. K-means clustering identified two groups of women with distinct phthalate/replacement biomarker profiles K-means clustering is useful for identifying subgroups of pregnant women with distinct patterns of biomarker concentrations. A major strength of this approach is that the identified clusters of women can be used in future studies to evaluate the relationships between population exposure patterns and adverse health outcomes. In our study, we identified two groups of women: those who had low phthalate (including ∑DEHTP) and those who had high phthalate (including ∑DEHTP) biomarker concentrations. A study of pregnant women from Ohio used k-means clustering to characterize patterns of exposure to numerous chemical classes, including phthalates, and identified three clusters of women that had different phthalate biomarker concentration patterns than those identified in our study (95). However, somewhat consistent with our results, another study of pregnant women from Wuhan, China evaluated trimester-specific population profiles of multiple chemical classes and observed that in any one trimester, groups of pregnant women had either high or low phthalate biomarker concentrations (97). This highlights a limitation of k-means because identified biomarker concentration patterns are population- specific and may not be generalizable to other cohorts. Additionally, k-means analyses that account for multiple classes of chemicals may yield different conclusions than those that focus on one chemical class. Therefore, additional studies using these approaches are needed to determine whether these patterns persist in other populations. Nevertheless, in our relatively homogenous population, k-means identified two relatively even clusters of women with concentrations of all phthalate biomarkers (including DEHTP) that were consistently higher or lower than the sample median. Interestingly, our 38 results also suggests that the plasticizer replacement biomarker ∑DiNCH was not related to high vs. low chemical cluster membership. In other words, DiNCH was uniformly distributed in our population. Because DiNCH was developed specifically for use in so- called sensitive applications, including medical tubing and children’s toys, it is possible that sources of DiNCH are less avoidable in our population than other phthalates/replacements. 2.6.2. Several maternal characteristics were important predictors of k-means clusters Although previous studies evaluated predictors of phthalate/replacement biomarker concentrations, most studies did not account for exposures to chemical mixtures. Therefore, we paired k-means clustering with logistic regression to identify characteristics associated with the two identified clusters of women. Overall, we observed that age, marital status, annual household income, parity, pre-pregnancy BMI, caffeine intake, enrollment year, and conception season remained the strongest important independent predictors of phthalate/replacement biomarker concentrations. Our findings that pregnant women who are younger, unmarried, have lower incomes, and had pre-pregnancy overweight/obesity have higher phthalate biomarker concentrations are in line with findings from several previous studies of pregnant women from the contiguous U.S. and Puerto Rico, Canada, Mexico, Europe, and China (18, 33, 90, 92, 93, 119-121). To our knowledge, this is the first study to identify caffeine consumption as a predictor of phthalate biomarker concentrations in pregnant women, although recent coffee consumption was associated with higher MCPP biomarker concentrations in an 39 adolescent population (122) and higher MEHHP-to-MECPP and MEOHP-to-MECPP ratios in U.S adults from NHANES 2001 – 2012 survey cycles (123). Whether this relationship is confounded by lifestyle factors that track with caffeine consumption in pregnancy, or is due to the diuretic nature of caffeine, or contamination by phthalates in food packaging used to prepare or consume caffeinated beverages will need to be further investigated. Our findings that women with at least one child have higher phthalate concentrations are consistent with some prior studies, but the literature related to phthalates and parity is generally mixed (92, 93, 124, 125). We also observed that women who conceived in the spring/summer/fall had lower odds of high phthalate/∑DEHTP. Given that women who conceived in the spring/summer provided most of their urine samples in fall/winter, our findings are somewhat consistent with studies of Swedish and Chinese pregnant women reporting higher phthalate biomarker concentrations in urine samples collected during spring/summer than fall/winter (88, 114). These studies suggest that these trends are likely due to seasonal differences in diet or personal care product use. Our findings related to enrollment period are supported by those from NHANES and other U.S. biomonitoring studies showing that phthalate biomarker concentrations are decreasing, especially phthalate plasticizer biomarker concentrations, while DiNCH and DEHTP metabolite concentrations are increasing over time (64, 65). These trends have also been confirmed in studies that include more recent sampling periods (126, 127) and suggest that women may be choosing to change their product use over time or that the use of phthalates/replacements in consumer products may be changing. 40 2.6.3. PCA identified four components of phthalate/replacement biomarker concentrations PCA is a dimension reduction method useful for identifying distinct, uncorrelated patterns of biomarker concentrations among highly correlated chemicals. Consistent with previous human biomonitoring studies (128), we observed that phthalate/replacement biomarkers were correlated along exposure sources, as the four identified components were heavily loaded by phthalates from plastics (∑DiNP or ∑DiNP2, MCNP, and MCPP), other phthalates (MBzP, ∑DBP, and ∑DiBP), major plasticizer phthalate and its replacements (∑DEHP, ∑DiNCH, and ∑DEHTP), and a major personal care product-related phthalate metabolite, MEP. While we only assessed phthalates, previous studies included additional chemical classes to identify broader exposure patterns during pregnancy (94- 97, 129). Using PCA, a few studies in pregnant populations from the USA (Ohio), Canada, and Europe reported that MEP always represented a unique component (94, 95, 129), which is consistent with our findings. In most populations, including ours, pregnant women have highest MEP concentrations relative to other phthalate metabolites likely due to their use of fragranced/perfumed products or cosmetics that contain diethyl phthalate, MEP’s parent compound (130, 131). Though a major limitation of PCA is that the identified components are unique to each population (similar to k-means), these consistent MEP findings in predominately white Western populations suggest that there may be certain exposure patterns that are consistent in pregnant women, although this needs to be further corroborated in non-Western populations. Most importantly, PCA in our study also identified a separate component heavily weighted for biomarkers of plasticizer replacements, DiNCH and DEHTP. This suggests that the replacements either have 41 shared exposure sources or exist together through behaviors aimed at reducing exposure to phthalates. These results are concerning since phthalate replacement concentrations are increasing over time (64), and their health impacts during pregnancy are poorly understood (132, 133). Additional studies in more diverse populations are needed to confirm these exposure biomarker patterns. 2.6.4. Several maternal characteristics were important predictors of PCs Pairing PCA results with linear regression helped us identify maternal characteristics that tracked most with the four identified exposure biomarker profiles/clusters. Similar to our results from k-means clustering, this approach also identified age, marital status, income, parity, pre-pregnancy BMI, caffeine intake, enrollment year, and conception season as important independent predictors of phthalate/replacement biomarker concentrations. However, these analyses additionally identified race/ethnicity, education, employment status, pregnancy intention, smoking status, alcohol intake, and diet quality as determinants of at least one PC. Our findings pertaining to sociodemographic characteristics are consistent with previous studies from the contiguous U.S. and Puerto Rico showing that women who are younger, have lower income, and are unmarried have higher “other” phthalate metabolite concentrations (MBzP, ∑DBP and ∑DiBP), and that women who are older, have higher incomes, and are employed have higher phthalate replacement biomarker concentrations (17, 18, 119, 134). Our findings also confirmed that black women, those who were older, and those who smoked during pregnancy have higher MEP concentrations, those with lower educational attainment have lower “other” phthalate biomarker concentration, and that women who had overweight or obesity before 42 pregnancy had higher phthalate plasticizer biomarker concentrations (18, 90, 91, 119, 121, 124, 134). However, inconsistent with results the Puerto Rican study (113), we observed that women with pre-pregnancy overweight or obesity had higher phthalate replacement biomarker concentrations. Urinary DEHTP metabolite concentrations in our population are at least two times higher than those in the Puerto Rico cohort, which may explain discrepancies in findings. Additionally, our findings regarding pregnancy intention are somewhat in line with those from a multi-center cohort of U.S. pregnant women finding that women with unplanned pregnancies generally had higher MCPP and lower MEP first trimester biomarker concentrations (115). We also observed that women with lower diet quality had lower “other” phthalate PC scores (strongest correlations with MBzP, ∑DBP and ∑DiBP). One possible explanation is that women with lower diet quality may be less likely to take supplements (135), which are a source of DBP (136). Alternatively, it is possible that our diet quality measure does not account for the use of plastic food storage containers or the use of plastic water bottles – dietary habits that have been associated with urinary BBzP and DBP metabolite biomarker concentrations (137). With regard to other studies using PCA, our findings are somewhat in line with those in pregnant women from Ohio, Canada, and China (discussed above). Race/ethnicity, household income, educational attainment, parity, pre-pregnancy BMI and diet were important predictors of phthalate biomarker PCs in the Ohio cohort (95), parity and smoking status were predictors of phthalate biomarker PCs in the Canadian cohort (94), while income, pre-pregnancy BMI, and employment status were important predictors of phthalate biomarkers PCs in the China cohort (97). However, two of these studies also 43 reported that birthplace was associated with phthalate biomarker PCs (94, 95), which was not evaluated in our study. Conversely, the multi-cohort European study found no associations of education or employment status with phthalate biomarker PCs (96). Given that only a few studies paired PCA with linear regression models, additional studies are needed to corroborate our findings. Nevertheless, our results show that phthalate/replacement exposure biomarker patterns in pregnant women are associated with sociodemographic-related lifestyle characteristics that may predict consumption of products containing phthalates/replacements. 2.6.5. Strengths and limitations Our study has several strengths. First, we systematically assessed a large number of a priori hypothesized maternal characteristics, some of which are not extensively studied in the literature (e.g. caffeine consumption, pregnancy intention, overall diet quality), providing novel information about predictors of phthalate/replacement biomarker concentrations in pregnancy. Second, we focused only on phthalates and their replacements to better understand biomarker profiles and predictors of this large class of endocrine disrupting chemicals. Additionally, we assessed associations of predictors with phthalate/replacement metabolites using mixture methods. It has been demonstrated that the use of mixture methods to evaluate association of phthalate biomarkers with preterm birth identified preterm birth risk beyond what was expected based on single-pollutant models. Our approach may help us better identify potential confounding factors of associations between cumulative phthalate/replacement exposure and pregnancy- related health outcomes (138). Third, phthalate/replacement biomarker concentrations 44 were quantified in pooled samples of up five first-morning urines, which is considered best practice for the assessment of non-persistent chemicals (139-141). Additionally, most women (94%) contributed all five urine samples, indicating that chemical biomarker concentrations measured in pooled samples in our population likely represent exposure across gestation. Lastly, our study is one of few to evaluate maternal predictors of DiNCH and DEHTP metabolite concentrations – plasticizer replacements to which pregnant women are becoming increasingly exposed. Interestingly, women in our study had higher DEHTP concentrations than women from NHANES, but also higher concentration than those reported by other pregnancy cohorts (121, 134). We are either capturing more recent trends or I-KIDS women (who represent women with higher socioeconomic status than the representative sample in NHANES) may be choosing products labeled as “phthalate free” that may contain replacements. This will need to be investigated in future studies. However, our study also has limitations. First, data on three urinary metabolites (DiNP metabolite MONP and DEHTP metabolites MEHHTP and MECPTP) were not available from our earliest participants, which limits our available sample size for assessing predictors of ∑DiNP (with the additional MONP metabolite) and ∑DEHTP, as well as ∑DiNCH that had lower detection frequencies compared to the other phthalate/replacement metabolites evaluated in our study. The loss of power and introduction of additional metabolites in multivariable analyses may account for differences in associations in the full versus sub-samples. Second, the timing for assessing early pregnancy lifestyle factors, which may change across a pregnancy, 45 relative to phthalate/replacement biomarker concentrations analyzed from a pooled sample to represent total gestational concentrations, may not be appropriate given the non-persistent nature of these chemicals. Third, the I-KIDS population is primarily comprised of non-Hispanic white women of high socioeconomic status, which may limit the generalizability of our findings to more diverse populations and our capacity to use refined categories for our sociodemographic variables. For example, our Asian category falsely homogenizes a diverse set of ethnicities, so additional studies may be needed to expand on this category. Lastly, although k-means clustering and PCA can be informative, the number and types of identified k-means clusters or PCs varies by population, making it challenging to use these data to generate universal recommendations for reducing prenatal phthalate exposure. 2.7. CONCLUSION In this midwestern U.S. population, we identified two distinct groups of pregnant women with specific phthalate/replacement biomarker concentration profiles, and four uncorrelated profiles of phthalate/replacement biomarkers that likely track together due to shared exposure sources. We also observed that several sociodemographic characteristics, early pregnancy lifestyle characteristics, enrollment year, and seasonality were associated with biomarker concentration profiles identified in k-means clustering and PCA. These findings contribute to the growing body of literature reporting confounding factors that should be considered in statistical models evaluating associations of biomarkers of phthalate/replacement mixtures with pregnancy-related health outcomes. Additionally, pairing unsupervised pattern identification methods like k- 46 means clustering or PCA with analyses evaluating predictors of phthalate/replacement biomarker concentration patterns is valuable for making recommendations targeted at limiting women’s consumption/use of phthalate-containing products during pregnancy. Future studies in more diverse populations are needed to confirm our findings (especially the magnitude of associations for each predictor) and to continue evaluating replacements such as DEHTP and DiNCH to fill in knowledge gaps about the determinants of these supposedly safer alternatives during and beyond pregnancy. 47 CHAPTER 3: MATERNAL PHTHALATE AND PHTHALATE ALTERNATIVE METABOLITES AND URINARY BIOMARKERS OF ESTROGENS AND TESTOSTERONES ACROSS PREGNANCY This article/chapter has been published in Environment International; Volume 155; Pacyga DC, Gardiner JC, Flaws JA, Li Z, Calafat AM, Korrick SA, Schantz SL, Strakovsky RS; Maternal phthalate and phthalate alternative metabolites and urinary biomarkers of estrogens and testosterones across pregnancy; Copyright Elsevier (2021); https://doi.org/10.1016/j.envint.2021.106676. 3.1. ABSTRACT Pregnant women are ubiquitously exposed to phthalates from food packaging materials and personal care products. Phthalates alter estrogen and testosterone concentrations in experimental models, but their ability to impact these hormones in human pregnancy is not well characterized. We recruited women ages 18-40 into the Illinois Kids Development Study (I-KIDS) in early pregnancy. Participants provided up to 5 first-morning urine samples across pregnancy (8-40 weeks gestation) that we pooled for quantification of 19 phthalate or phthalate alternative metabolites. Either individual (ng/mL) or molar sums (nmol/mL) of metabolites were used as exposure biomarkers. We summed urinary concentrations (ng/mL) of eight major estrogen (SumEstrogens) and two major testosterone (SumTestosterones) metabolites measured at median 13, 28, and 34 weeks gestation. We also estimated the ratio of estrogens-to-androgens. Linear mixed-effects models assessed relationships of phthalates/alternatives as continuous measures or as concentration quartiles with SumEstrogens, SumTestosterones, and the Estrogen/Androgen ratio in 434 women. In our models, we controlled for age, race, 48 education, parity, smoking in the first trimester, pre-pregnancy body mass index, diet quality, conception season, fetal sex, and gestational age at hormone assessment. We also explored whether gestational age at hormone assessment or fetal sex modified these associations. All biomarkers and outcomes were specific gravity-adjusted, and continuous exposures and outcomes were also natural log-transformed. Most participants were non-Hispanic white (80.9%), college educated (82.2%), and had urinary phthalate/alternative metabolite concentrations similar to those of reproductive-aged U.S. women. Overall, select phthalate metabolites were positively associated with SumEstrogens and SumTestosterones, but negatively associated with the Estrogen/Androgen ratio. For example, SumEstrogens was 5.1% (95%CI: 1.8, 8.5) higher with every 2-fold increase in sum of di(2-ethylhexyl) phthalate metabolites, while SumTestosterones was 7.9% (95%CI: 1.0, 15.3) higher and Estrogen/Androgen ratio was -7.7% (95%CI: -13.6, -1.4) lower with every 2-fold increase in monoethyl phthalate. However, phthalate alternatives were only positively associated with SumEstrogens, which was 2.4% (95%CI: 0.4, 4.5) and 3.2% (95%CI: 0.7, 5.8) higher with every 2-fold increase in sum of di(isononyl) cyclohexane-1,2-dicarboxylate metabolites and sum of di(2-ethylhexyl) terephthalate metabolites, respectively. Gestational age- and fetal sex- specific associations were only consistently observed for associations of phthalates/alternatives with SumEstrogens, where associations were strongest in mid-to- late pregnancy in women carrying females. Phthalates/alternatives may impact gestational hormones, with potential for gestational age- and fetal sex-specific associations. Whether maternal urinary estrogens and testosterones mediate 49 associations of phthalates/alternatives with pregnancy and fetal outcomes merits further investigation. 3.2. KEYWORDS Phthalate replacements, phthalates, pregnancy, urinary estrogen metabolites, urinary testosterone metabolites. 3.3. INTRODUCTION Estrogens and androgens (e.g. testosterone) are sex-steroid hormones that support pregnancy progression and fetal development (142). While maternal ovaries and adrenal glands contribute to maternal circulating sex-steroid hormones in early pregnancy, by the second trimester, the maternal-fetal-placental unit is almost solely responsible for synthesizing sex-steroid hormones from maternal and fetal cholesterol (4, 143). Longitudinal studies demonstrate that estrogens and testosterone increase across gestation to support numerous pregnancy-related events (6, 144). For example, substantial increases in major circulating estrogens (i.e. estrone, estradiol, estriol) are responsible for supporting implantation, placental angiogenesis, maternal metabolism, and parturition (6). Although less understood, minor increases in gestational testosterone support cervical remodeling and myometrial function responsible for parturition (7). As a result, deviations from normal gestational estrogen and testosterone patterns may result in adverse maternal and fetal outcomes. To that end, observational studies suggest that disrupted estrogens and androgens are associated with increased risk of pre-eclampsia, gestational diabetes, and pre-term birth (7-9). To protect maternal and child health, it is 50 critical to understand factors that could disrupt the patterns of estrogens and testosterone in pregnancy. Pregnant women are widely exposed to phthalates, which are endocrine disrupting chemicals that modify the synthesis and function of sex-steroid hormones. In the 2015- 16 National Health and Nutrition Examination Survey (NHANES), >90% of reproductive- aged 18-40 year old U.S. women had detectable concentrations of at least one phthalate metabolite in their urine, which may be related to the use of phthalates in common consumer products, including food packaging materials and personal care products (145, 146). Prenatal phthalate exposure has been associated with poor pregnancy and birth outcomes, including early pregnancy loss and pre-term birth (24, 29), as well as adverse child outcomes, including cognitive and metabolic problems (45, 50). Evidence for the endocrine disrupting potential of phthalates originates from in vitro and in vivo studies showing that certain phthalates can bind to estrogen and androgen receptors and may have weak estrogenic or anti-androgenic properties (19-22). Phthalates may also indirectly impact estrogen and testosterone synthesis and function by altering follicle stimulating hormone concentrations or by interacting with peroxisome proliferator- activated receptors or thyroid receptors, which are part of estrogen and androgen regulatory pathways (147). Consistently, studies in animal models and humans have also demonstrated adverse associations of prenatal phthalate exposure with reproductive capacity or sex hormone-mediated outcomes (148, 149). However, few epidemiological studies have directly evaluated the impact of prenatal phthalate exposure on maternal hormone levels during pregnancy (77, 150-152). Additionally, while policies have led to 51 reduced use of phthalates in certain consumer products (153, 154), phthalate alternatives have been introduced to replace phthalates in some plastic materials, with limited data related to the consequences of exposure to these replacements on human health. To our knowledge, five prospective cohort studies have evaluated associations of maternal phthalate exposure with serum or plasma sex-steroid hormones in pregnant women (77, 150-152, 155), and only one evaluated relationships of phthalate alternatives with these hormones (152). Furthermore, only two studies evaluated fetal sex-specific associations, which may be important given that maternal phthalate exposures are sex- specifically associated with fetal and child health outcomes (151, 155). Additionally, previous studies related to phthalates and maternal hormones only evaluated associations of phthalates with sex-steroid hormones during a single trimester – either in the first trimester (150) or the second/third trimesters (77, 151, 152), although a recent study evaluated maternal sex-steroid hormones in the first trimester and at term upon arrival to the hospital for delivery (155). Given the dynamic cross-pregnancy changes in estrogen and, to a lesser extent, testosterone levels, and the fact that previous studies collectively suggest that associations of phthalates/alternatives with hormones differ across pregnancy (77, 150-152, 155), it is critical to evaluate these associations at multiple timepoints across gestation. Repeated blood sampling across gestation may not be feasible in large cohorts of pregnant women, but sex-steroid hormones can be measured in urine (143), which in some cases may be easier to obtain than blood. This approach was developed in non- 52 pregnant women as a proxy biomarker of circulating sex-steroid hormones (143, 156). More importantly, there is some evidence that urinary hormones may recapitulate observed associations of plasma/serum hormones with certain reproductive and lifestyle factors (157, 158). For example, urinary estrogen metabolites correlate with age at first birth and smoking status analogously to correlates of serum estrogen (157, 158), and urinary testosterone metabolite concentrations decrease with women’s age similar to what is observed with serum testosterone (159). Conversely, other studies suggest that urinary hormones or their metabolites may not reflect circulating hormone concentrations, but rather represent hormone metabolism (160, 161). To circumvent these discrepancies with assessing hormones in blood versus urine, previous studies created sums of multiple urinary estrogen and testosterone metabolites to provide an estimate of total circulating estrogen or testosterone concentrations (160, 161). While this approach does not allow for evaluating associations between phthalates/alternatives and individual hormones, it does make it easier to conduct cross-pregnancy assessment of hormone status in large pregnancy cohorts. Given the importance of estrogens and testosterone in pregnancy, our first objective was to evaluate associations of gestational urinary biomarkers of phthalates/alternatives exposures with urinary estrogens, testosterones, and the estrogen/androgen ratio. Unlike previous studies, we evaluated hormones across gestation in urine and explored whether these relationships are dose-dependent. Because estrogens/androgens change across pregnancy and some differ by fetal sex, our second objective was to evaluate whether associations of gestational phthalate/alternative metabolite concentrations with urinary 53 estrogens/testosterones differ across three gestational timepoints, or if they differ depending on the sex of the fetus. Results from our analyses provide additional insights into the endocrine disrupting potential of phthalates and phthalate alternatives across pregnancy, which may have long-term implications for maternal and child health. 3.4. MATERIALS AND METHODS 3.4.1. Illinois Kids Development Study (I-KIDS) recruitment and enrollment This study includes pregnant participants from I-KIDS, an ongoing prospective pregnancy cohort with the overarching goal of evaluating the impacts of prenatal environmental chemical exposures on infant neurodevelopment. We recruited pregnant women from two local obstetric clinics in Champaign-Urbana, IL at their first prenatal care appointment. Women who expressed interest in the study were contacted by I-KIDS staff and were eligible to participate if they were ≥10 but <15 weeks pregnant, 18-40 years old, fluent in English, not in a high-risk pregnancy or carrying multiple fetuses, living within a 30-minute drive of the University of Illinois campus, and not planning to move out of the area before their child’s first birthday. The current study includes the first 439 women who enrolled in I-KIDS between December 2013 and February 2018. For the current study (and all chemical/hormone analyses) women must have remained in the study through the birth of their infant. Enrolled women provided written informed consent according to the Institutional Review Board at the University of Illinois. The analysis of de-identified specimens at the Centers for Disease Control and Prevention (CDC) laboratory was determined not to constitute engagement in human subjects research. 54 3.4.2. Collection of maternal sociodemographic and lifestyle information at enrollment Immediately after enrollment, we visited each pregnant I-KIDS participant at home to obtain information about their demographics, lifestyle, and health. We interviewed participants approximately monthly after this initial visit to ascertain important pregnancy- related updates, but only used information collected at the initial visit in the current analysis. Specifically, the following relevant demographic and lifestyle variables were self- reported by women at baseline: maternal age, race/ethnicity, highest educational level attained, smoking status since conception, and parity. Women also reported pre- pregnancy weight in pounds and height in feet and inches, which we used to calculate pre-pregnancy body mass index (BMI in kg/m2). Self-reported pre-pregnancy BMI has been shown to be highly correlated with measured first trimester BMI (98-100), which we also confirmed in a subset of women from our study (r = 0.93; data not shown). Estimated due date based on the last menstrual period was collected at baseline and confirmed after the first trimester ultrasound, while information about fetal sex was collected at birth. At 8-15 and 32-40 weeks gestation, women completed a semi-quantitative food frequency questionnaire (FFQ) adapted from the full-length Block-98 FFQ (NutritionQuest, Berkeley, CA). The FFQ asked about maternal diet during the previous three months (101) and was used to calculate the Alternative Healthy Eating Index 2010 (AHEI-2010) in early and late gestation. AHEI-2010 is an 11 component diet quality measure (scored out of 110) based on foods/nutrients predictive of chronic disease risk, where a higher score is reflective of better overall diet quality (102). Because overall diet quality was relatively stable from early to late pregnancy in the I-KIDS population (data not shown), we used the mean of 55 AHEI-2010 scores at the two timepoints to estimate maternal diet quality across gestation. 3.4.3. Collection and processing of urine samples for chemical and hormone analyses Pregnant women provided first-morning urine samples at 8-15, 14-22, 19-28, 25-33, and 32-40 weeks gestation (median 13, 17, 23, 28, and 34 weeks, respectively). Urine samples were collected in polypropylene urine cups. All samples were refrigerated immediately after collection and transported on ice to the I-KIDS laboratory. Within 24 hours of collection, urine samples were warmed for 30 minutes to room temperature, vortexed, and assessed for specific gravity using a handheld refractometer (TS400; Reichert Technologies, Depew, NY). Each urine sample was aliquoted into polypropylene cryovial tubes (Nalgene, Rochester, NY) using disposable polyethylene bulb transfer pipettes (Fisher Scientific, Ann Arbor, MI). Duplicates and purified water blanks were collected and analyzed for every 10 samples. In addition to creating individual aliquots of urines at each timepoint, we also pooled all five urines for each participant by adding 900 µL of fresh urine from each timepoint to a 5 mL cryovial tube containing frozen urine from previous gestational timepoints. Specific gravity of pooled samples was measured at the end of pregnancy, when each pooled sample was thawed and vortexed. All urine was stored at -80 °C. 3.4.4. Quantification of urinary phthalate/alternative metabolites Because phthalates/phthalate alternatives have short half-lives and high within-individual exposure variability (162), the current study assessed phthalate and phthalate alternative metabolites in pooled samples of up to five first morning urines to approximate maternal 56 phthalate/alternative exposure across pregnancy. We shipped frozen pooled urine samples to the CDC on dry ice in three batches in the order of participant enrollment (batch one enrolled December 2013-February 2015, batch two enrolled February 2015- July 2016, and batch three enrolled July 2016-February 2018). Urinary phthalate and phthalate alternative metabolite concentrations were quantified at the CDC using on-line solid phase extraction coupled with isotope dilution-high performance liquid chromatography-electrospray ionization-tandem mass spectrometry (104). The following metabolites were quantified in all batches: mono(2-ethylhexyl) phthalate (MEHP), mono(2-ethyl-5-hydroxyhexyl) phthalate (MEHHP), mono(2-ethyl-5-oxohexyl) phthalate (MEOHP), mono(2-ethyl-5-carboxypentyl) phthalate (MECPP), mono-isononyl phthalate (MiNP), monocarboxyoctyl phthalate (MCOP), monocarboxynonyl phthalate (MCNP), mono(3-carboxypropyl) phthalate (MCPP), monobenzyl phthalate (MBzP), monoethyl phthalate (MEP), mono-n-butyl phthalate (MBP), mono-hydroxybutyl phthalate (MHBP), mono-isobutyl phthalate (MiBP), mono-hydroxy-isobutyl phthalate (MiHBP), cyclohexane-1,2-dicarboxylic acid-monohydroxy isononyl ester (MHiNCH), and cyclohexane-1,2-dicarboxylic acid-mono(carboxyoctyl) ester (MCOCH). Three additional metabolites were added to the CDC analytical panel for women in batches two and three (monooxononyl phthalate (MONP), mono(2-ethyl-5-hydroxyhexyl) terephthalate (MEHHTP), and mono(2-ethyl-5-carboxypentyl) terephthalate (MECPTP)). 3.4.5. Measurement of urinary estrogen and testosterone metabolites We measured eight major estrogens (estrone, estradiol, estriol, 16α-hydroxyestrone, 2- hydroxyestrone, 2-methoxyestrone, 4-hydroxyestrone, and 4-methoxyestrone) and two 57 major testosterones (testosterone and 5α-dihydrotestosterone) in three individual first- morning urine samples collected at median 13, 28, and 34 weeks gestation, corresponding to early, middle, and late gestational plasma estrogen and testosterone concentrations (metabolites are bolded in Figure 5, adapted from: (4, 143)). All samples were analyzed in one analytical batch using methods adapted from Xu et al. (143). Urine samples for hormone analyses were prepared as follows: 500 µL of urine was mixed with 100 µL 1 M acetate buffer (pH 4.0), 20 µL of 0.5 µg/mL D3-testosterone (Sigma Aldrich Co. St. Louis, MO), and 20 µL beta-glucuronidase (Roche through Sigma Aldrich). The mixture was then vortexed, incubated in a heat block at 63 °C for 30 minutes, and centrifuged for 5 minutes at 8,000 rpm. Standards for all hormones were purchased from Steraloids Inc, Newport, RI. Solid-phase extraction (SPE) cleanup was performed using polymeric reverse phase cartridges (StrataTM-X, Phenomenex, Torrence, CA). Prepared and cleaned samples were analyzed with a 5500 QTRAP LC/MS/MS system (AB Sciex, Framingham, MA) with 1200 series HPLC system (Agilent Technologies) in the Roy J. Carver Biotechnology Center Metabolomics Lab at the University of Illinois at Urbana- Champaign. 58 Figure 5. Estrogen and testosterone metabolism pathway. Maternal estrogens and testosterones are cholesterol-derived hormones that are primarily synthesized from maternal cholesterol in the ovaries, adrenal glands, and adipose tissue (to a lesser extent) in the first trimester. The maternal-fetal-placental unit becomes the primary source of hormones after the first trimester. Estrogens and testosterones are metabolized in the maternal liver, and the parent compounds as well as resulting metabolites are excreted in urine. The bolded compounds are the eight major urinary estrogen and two major urinary testosterone parent compounds or metabolites measured in this study. DHEA, dihydroepiandrostenediene; DHEAS, dihydroepiandrostenediene sulfate. 59 3.4.6. Statistical analysis 3.4.6.1. Final sample size and covariate selection A total of 439 women were available for statistical analyses. In all statistical models, we included the following covariates selected a priori using previous literature: maternal age, race/ethnicity, education, parity, smoking since conception, pre-pregnancy BMI, AHEI- 2010, season of conception, fetal sex, and timepoint of hormone assessment. We evaluated correlations between all potential confounders to test for multicollinearity. Three women had missing information about pre-pregnancy BMI, and two others had missing information about race/ethnicity or AHEI-2010 (Table 7). Therefore, 434 pregnant women were included in final statistical models. For hormone analysis, a total of 433, 424, and 426 women contributed urine samples at median 13, 28, and/or 34 weeks gestation, respectively, with 414 contributing urine samples at all three timepoints and 20 contributing urine samples at only two timepoints. Maternal age, AHEI-2010, and pre- pregnancy BMI were included as continuous variables, while the remaining covariates were categorical (Table 7). 3.4.6.2. Exposure and outcome variables To avoid bias associated with imputing values below the limit of detection (LOD) (107), we used instrument-read values for all samples. Across the chemical and hormone analyses, only 5 values were zero (n=1 for SumDiNCH; n=3 and n=1 for SumTestosterones at median 13 and 28 weeks gestation, respectively), so in final statistical models we added a constant 0.0001 to these zero values before natural log- transformation (ln-transformation) to avoid undefined estimates (108). To account for 60 urine dilution, we used the following formula to adjust all urinary chemical and hormone metabolite concentrations: Pc = P[(SG − 1)/(SGi − 1)], where Pc is the specific gravity adjusted chemical or hormone metabolite concentration, P is the measured chemical or hormone metabolite concentration (ng/mL), SG is the median specific gravity of the pooled samples used for chemical analysis (1.016) and three urine samples used for hormone analysis (1.015), and SGi is the specific gravity of each individual urine sample (109). We approximated exposure to phthalate and phthalate alternative parent compounds that are metabolized and excreted as multiple urinary metabolites using the following molar- sum (in nmol/mL) equations: sum of di(2-ethylhexyl) phthalate metabolites (SumDEHP) = (MEHP/278) + (MEHHP/294) + (MEOHP/292) + (MECPP/308), sum of di-isononyl phthalate metabolites (SumDiNP) = (MiNP/292) + (MCOP/322), sum of di-n-butyl phthalate metabolites (SumDBP) = (MBP/222) + (MHBP/238), sum of di-iso-butyl phthalate metabolites (SumDiBP) = (MiBP/222) + (MHiBP/238), sum of di(isononyl) cyclohexane-1,2-dicarboxylate metabolites (SumDiNCH) = (MHiNCH/314) + (MCOCH/328), and sum of di(2-ethylhexyl) terephthalate metabolites (SumDEHTP) = (MEHHTP/294) + (MECPTP/308). We excluded MONP from SumDiNP because this metabolite was only measured in batches 2 and 3, and we did not observe marked differences in associations between SumDiNP and hormones when MONP was included in the sum (data not shown). We approximated exposure to di-isodecyl phthalate, di-n- octyl phthalate, benzylbutyl phthalate, and diethyl phthalate using the concentrations (in ng/mL) of their non-molar converted major urinary metabolites MCNP, MCPP, MBzP, and 61 MEP, respectively. Additionally, we molar-summed (nmol/mL) some phthalate metabolites based on common exposure sources by calculating the sum of phthalate metabolites of parent compounds found in plastics (SumPlastics) and personal care products (SumPCP) as follows (163): SumPlastics = (MEHP/278) + (MEHHP/294) + (MEOHP/292) + (MECPP/308) + (MiNP/292) + (MCOP/322) + (MCNP/336) + (MCPP/252) + (MBzP/256) and SumPCP = (MEP/194) + (MBP/222) + (MHBP/238) + (MiBP/222) + (MHiBP/238). SumPlastics and SumPCP are reflective of high and low molecular weight phthalate metabolites, respectively. Urine and blood likely capture different hormone forms, such that unconjugated (biologically active) hormones are measured in blood, whereas conjugated (biologically inactive) hormones are generally measured in urine (161). Therefore, individual urinary hormones, especially the downstream metabolites of the parent hormones, may not be representative of hormones in circulation, but rather provide insights into steroid hormone metabolism. This is particularly evident when comparing studies that quantified maternal parent estrogen concentrations (estrone, estradiol, and estriol) across pregnancy in blood versus urine, where estradiol and estriol tend to have the highest concentrations in blood and urine, respectively, relative to the other parent estrogens (4, 164). To circumvent this, we created hormone sums at each of the three gestational timepoints to represent total estrogens (SumEstrogens of eight major urinary estrogen metabolites) and testosterones (SumTestosterones of two major testosterone metabolites) in early, mid, and late gestation – an approach that has been previously used to characterize hormone status using hormones measured in urine (160, 161). Because the estrogen-to-androgen ratio 62 may be an important indicator of pregnancy health (165), we additionally created an Estrogen/Androgen ratio by dividing SumEstrogens by SumTestosterones at each gestational timepoint. The Estrogen/Androgen ratio was only created for women who had non-zero values for both SumEstrogens and SumTestosterones. 3.4.6.3. Linear mixed model approach We used linear mixed models to accommodate our longitudinal prospective design with hormone outcomes at three timepoints in gestation. SumEstrogens, SumTestosterones, and Estrogen/Androgen ratio were analyzed separately. An unstructured covariance 3×3 matrix was specified for the model’s residuals. All statistical analyses were conducted in SAS Software, version 9.4 (SAS Institute Inc, Cary, NC). Associations were considered significant at P < 0.05, and we did not adjust for multiple comparisons (166). The first part of our first objective was to evaluate associations of continuous phthalate/alternative biomarkers with maternal sums of urinary hormone metabolites across pregnancy. We ln-transformed our exposures and outcomes to fit normality assumptions and added 0.0001 to phthalate/alternative or hormones that had zero as a minimum value (see Section 3.4.6.2 for specific details). The second part of our first objective was to assess potential dose-response relationships between phthalates/alternatives and hormones. Phthalates were categorized based on quartiles of urinary phthalate/alternative biomarker concentrations, while hormones were ln- transformed as described previously. For our second objective, because previous studies suggest that associations of phthalate and phthalate alternative biomarkers with 63 hormones may be gestational age- or fetal sex-specific (77, 151, 152, 155), we also evaluated gestational age- and fetal sex-specific relationships between phthalate/alternative biomarkers and hormones by including a three-way interaction (and all relevant two-way interactions) between exposure, gestational age at hormone sampling, and fetal sex. We only evaluated gestational age- and fetal sex-specific associations of continuous phthalate/alternative biomarkers with hormones. 3.4.6.4. Reporting of results All gestational age- and fetal sex-specific results were reported regardless of the significance of the three-way interaction term. All β-estimates and 95% confidence intervals (CIs) were back transformed and presented as a percent change in hormones in tables and figures. In models with both ln-transformed phthalate/alternative biomarkers and hormones (objectives 1 and 2), β-estimates and 95% CIs were back transformed using the equation [((2.00)β – 1)*100] to represent a percent change in hormones for each 2-fold increase in phthalate/alternative (Table 8). For models with ln-transformed hormones and phthalate/alternatives biomarkers in quartiles (objective 1), β-estimates and 95% CIs were back transformed using the equation [(eβ – 1)*100] to represent the percent change in hormones among women in quartiles two (Q2), three (Q3), and four (Q4) of urinary phthalate/alternative biomarkers concentrations, which were each compared to quartile one (Q1) (Table 8, Figures 8-10). 64 3.5. RESULTS 3.5.1. Demographic and lifestyle characteristics of the I-KIDS population Baseline characteristics of 439 pregnant I-KIDS women are reported in Table 7. Participants had a median (range) age of 30 years (18-40), and the majority were non- Hispanic white (81%) and college educated (82%). Approximately 52% of women were nulliparous, while 48% were primiparous or multiparous. The majority of women did not smoke since conception (88%), 52% were normal weight and 45% were overweight or had obesity before pregnancy, and median (range) diet quality score measured by the AHEI-2010 was 55.8 (28.1-82.8) out of 110. Season of conception was equally distributed across all seasons, and fetal sex was also relatively evenly distributed, with 52% and 48% of women carrying a female or male fetus, respectively. 65 Table 7. Baseline demographic and lifestyle characteristics of I-KIDS pregnant women. Demographic or lifestyle characteristic I-KIDS women (n=439)1 median (range) or n (%) Age 30.0 (18.0 – 40.0) Alternative Healthy Eating Index-2010 (1 missing) 55.8 (28.1 – 82.8) Race/ethnicity (1 missing) Non-Hispanic White 355 (80.9) Others2 83 (18.9) Education Some college of less 78 (17.8) College graduate or higher 361 (82.2) Parity Nulliparous 228 (51.9) Primiparous 139 (31.7) Multiparous 72 (16.4) Smoking during 1st trimester No 385 (87.7) Yes 21 (4.8) Unknown 33 (7.5) Pre-pregnancy BMI (3 missing) Underweight (<18.5 kg/m2) 10 (2.3) Normal weight (18.5-24.9 kg/m2) 227 (51.7) Overweight (25-29.9 kg/m2) 103 (23.5) Obese (≥30 kg/m2) 96 (21.9) Season of conception Winter 108 (24.6) Spring 114 (25.9) Summer 100 (22.8) Fall 117 (26.7) Fetal sex Female 229 (52.2) Male 210 (47.8) 1Percentages may not add up to 100% due to missing.2Includes Hispanic Whites, Native American or Alaska Natives, Asians, Blacks or African Americans, Native Hawaiians or other Pacific Islanders, and others. BMI, body mass index; I-KIDS, Illinois Kids Development Study. 3.5.2. Urinary phthalate and phthalate alternative metabolite concentrations The concentrations of most phthalate and phthalate alternative metabolites were detectable in 100% of women, except for the following metabolites (% of women with metabolite concentrations >LOD): MEHP (74%), MiNP (42%), MCPP (97%), MBzP (99%), MHBP (90%), MiBP (99%), MHiBP (99%), MHiNCH (77%), and MCOCH (49%) (data not shown). Urinary phthalate/alternative metabolite concentrations in I-KIDS women were generally comparable to those in 18-40-year-old women from NHANES 66 cycles 2013-14 and 2015-16 (Figure 6). Figure 6. Urinary phthalate and phthalate alternative metabolite concentrations in pregnant women from I-KIDS 2013-18 compared to women from the NHANES 2013- 16. Urinary phthalate (A-J, O-S) and phthalate alternative (K-N) metabolite concentrations were assessed from pooled sample of five urine samples per participant (n=439). Results are presented as 1.5 times the interquartile range below and above the 25th and 75th percentiles (lower and upper endpoints of whisker), the 25th and 75th percentiles (lower and upper edges of box), median (line inside box), and mean (diamond). Urinary phthalate and phthalate alternative metabolite concentrations assessed from a spot urine sample per participant were obtained for 18-40-year-old reproductive-aged women from NHANES survey years 2013-14 (n=394 or 392) and 2015-16 (n=348). I-KIDS, Illinois Kids Development Study; NHANES, National Health and Nutrition Examination Survey. 67 3.5.3. Urinary sex-steroid hormone concentrations As expected, urinary SumEstrogens, SumTestosterones, and the Estrogen/Androgen ratio increased across gestation (Figure 7; Ptime < 0.0001). Median (range) urinary specific gravity-adjusted hormone concentrations at 8-15, 25-33, and 32-40 weeks gestation were as follows: SumEstrogens: 3064.4 ng/mL (313.5-57,879.9), 11,168.9 ng/mL (2,845.3-40,420.4), and 14350.1 ng/mL (4,063.1-41,561.0), respectively (Figure 7A); SumTestosterones: 3.6 ng/mL (0.0-61.9), 3.4 ng/mL (0.0-84.6), and 4.3 ng/mL (0.1- 73.2), respectively (Figure 7B); and Estrogen/Androgen ratio: 848.3 (75.5-236,684.0), 3,720.2 (43.3-332,635.5), and 3,659.1 (176.7-106,960.2), respectively (Figure 7C). 68 Figure 7. Distribution of urinary (A) SumEstrogen (ng/mL), (B) SumTestosterone (ng/mL), and (C) Estrogen/Androgen ratio at 8-15, 25-33, and 32-40 weeks gestation (n=439). Results are presented as 1.5 times the interquartile range below and above the 25th and 75th percentiles (lower and upper endpoints of whisker), the 25th and 75th percentiles (lower and upper edges of box), median (line inside box), and mean (diamond). Linear mixed models were used to assess whether hormone concentrations differed across the three gestational timepoints (Ptime). 3.5.4. Overall associations of phthalate/alternative biomarkers with hormones In linear models, SumDEHP, MCNP, MCPP, SumPlastics, SumDiNCH, SumDEHTP, and SumDBP were positively associated with SumEstrogens along with less precise positive associations observed for SumDiBP and SumPCP with SumEstrogens (Table 8). For example, each 2-fold increase in SumDEHP and SumDEHTP was associated with 7.2% (95%CI: 3.9, 10.6) and 3.7% (95%CI: 1.2, 6.2) increase in SumEstrogens, respectively. Most analyses where phthalate/alternative biomarkers were modeled in quartiles generally supported the linear models. SumEstrogens were higher in women at the upper 69 quartiles of SumDEHP (Q3, Q4), MCPP (Q2, Q3, Q4), SumPlastics (Q3, Q4), SumDiNCH (Q4), SumDBP (Q2, Q3, Q4), SumDiBP (Q2, Q3, Q4), and SumPCP (Q2, Q3, Q4), as well as at MEP Q3 compared to those in Q1. For example, compared to Q1, SumEstrogens was 8.5% (95%CI: -0.1, 18.0) and 18.2% (95%CI: 8.4, 28.8) higher in SumDEHP Q3 and Q4. Several phthalate and phthalate alternative biomarkers were also associated with urinary SumTestosterone concentrations across pregnancy (Table 8). In linear models, every 2- fold increase in MBzP and MEP was associated with 5.5% (95%CI: -1.1, 12.5) and 9.3% (95%CI: 1.8, 17.5) increase in SumTestosterones, respectively. In models where phthalate/alternative biomarkers were modeled as quartiles, compared to the lowest quartile, SumTestosterones were higher at higher quartiles of SumDEHP (Q3), MCPP (Q3), MBzP (Q4), SumPlastics (Q3), and MEP (Q3 and Q4). For example, SumTestosterones were 40.5% (95%CI: 7.4, 83.8) and 34.7% (95%CI: 2.7, 76.7) higher in MEP Q3 and Q4, respectively, compared to Q1. However, we additionally observed potential non-linear associations of SumDiBP and SumPCP, with SumTestosterones where strongest positive associations emerged at Q2 compared to the lowest quartile. Some phthalate, but not phthalate alternative, biomarkers were associated with the Estrogen/Androgen ratio (Table 8). Specifically, in linear models, the ratio tended to be negatively associated with SumDEHP and MBzP, and was negatively associated with MEP and SumPCP, such that each 2-fold increase in MEP and SumPCP was associated with -10.0% (95%CI: -16.4, -3.0) and -10.1% (95%CI: -18.8, -0.6) lower 70 Estrogen/Androgen ratio. In quartile analyses, compared to those in Q1, the Estrogen/Androgen ratio tended to be lower for women in SumDEHP Q3 and MCPP Q3, and was lower in women at higher quartiles of MEP (Q3 and Q4) and SumPCP (Q4). For example, the Estrogen/Androgen ratio was -29.2% (95%CI: -46.4, -6.5) and -30.3% (95%CI: -47.4, -7.7) lower in MEP Q3 and Q4, respectively, compared to Q1. 71 Table 8. Overall associations of phthalate and phthalate alternative biomarkers with SumEstrogens, SumTestosterones, and Estrogen/Androgen ratio. % Change (95%CI) in % Change (95%CI) in % Change (95%CI) in Biomarker P P P SumEstrogens SumTestosterones Estrogen/Androgen ratio SumDEHP Linear association 7.2 (3.9, 10.6) <.0001 -1.7 (-11.0, 8.5) 0.73 10.7 (-0.1, 22.7) 0.05 Q2 (ref=Q1) -2.6 (-10.4, 5.8) 0.53 17.5 (-9.9, 53.2) 0.23 -17.5 (-37.4, 8.6) 0.17 Q3 (ref=Q1) 8.5 (-0.1, 18.0) 0.05 42.6 (9.3, 86.0) 0.01 -26.5 (-44.2, -3.2) 0.03 Q4 (ref=Q1) 18.2 (8.4, 28.8) 0.0002 10.8 (-15.8, 45.7) 0.46 9.9 (-17.3, 45.9) 0.51 SumDiNP Linear association 1.1 (-2.1, 4.4) 0.51 1.3 (-8.5, 12.2) 0.80 -2.6 (-12.3, 8.2) 0.62 Q2 (ref=Q1) 3.1 (-5.2, 12.1) 0.48 -6.1 (-28.1, 22.5) 0.64 3.9 (-21.2, 37.0) 0.79 Q3 (ref=Q1) -2.6 (-10.8, 6.2) 0.55 7.8 (-18.2, 42.0) 0.59 -4.6 (-28.4, 27.1) 0.75 Q4 (ref=Q1) 8.3 (-0.9, 18.3) 0.08 3.3 (-21.9, 36.8) 0.82 7.5 (-19.7, 43.8) 0.63 MCNP Linear association 3.2 (0.4, 6.1) 0.02 5.8 (-3.1, 15.4) 0.21 -0.6 (-9.2, 8.9) 0.90 Q2 (ref=Q1) -0.5 (-8.8, 8.5) 0.90 15.6 (-12.0, 51.7) 0.30 -16.9 (-37.4, 10.3) 0.20 Q3 (ref=Q1) 2.7 (-5.8, 12.0) 0.54 20.9 (-7.9, 58.5) 0.17 -16.0 (-36.7, 11.3) 0.22 Q4 (ref=Q1) 6.2 (-2.7, 16.0) 0.18 21.8 (-7.5, 60.4) 0.16 -11.2 (-33.3, 18.2) 0.41 MCPP Linear association 4.4 (1.7, 7.1) 0.001 3.6 (-4.6, 12.5) 0.40 0.3 (-8.0, 9.3) 0.95 Q2 (ref=Q1) 9.8 (1.0, 19.5) 0.03 25.0 (-4.2, 63.2) 0.10 -13.6 (-34.5, 13.9) 0.30 Q3 (ref=Q1) 9.0 (0.3, 18.6) 0.04 36.7 (4.9, 78.0) 0.02 -23.9 (-42.2, 0.2) 0.05 Q4 (ref=Q1) 15.0 (5.7, 25.1) 0.001 20.0 (-8.0, 56.5) 0.18 -2.0 (-25.6, 29.2) 0.89 MBzP Linear association 1.4 (-0.7, 3.5) 0.19 5.5 (-1.1, 12.5) 0.11 -5.1 (-11.3, 1.5) 0.12 Q2 (ref=Q1) 6.4 (-2.1, 15.7) 0.15 5.4 (-18.9, 37.1) 0.69 -6.4 (-28.8, 23.0) 0.63 Q3 (ref=Q1) 7.1 (-1.8, 16.8) 0.12 12.8 (-14.2, 48.1) 0.39 -13.3 (-34.6, 15.1) 0.32 Q4 (ref=Q1) 6.9 (-1.9, 16.5) 0.13 19.7 (-8.5, 56.7) 0.19 -15.5 (-36.1, 11.8) 0.24 SumPlastics Linear association 5.4 (2.3, 8.7) 0.001 1.9 (-7.5, 12.2) 0.71 5.2 (-4.8, 16.3) 0.32 Q2 (ref=Q1) 5.4 (-3.2, 14.7) 0.23 -10.0 (-31.2, 17.6) 0.44 21.3 (-8.3, 60.4) 0.18 Q3 (ref=Q1) 15.6 (6.1, 26.0) 0.001 31.3 (0.1, 72.3) 0.05 -7.1 (-30.0, 23.3) 0.61 Q4 (ref=Q1) 14.1 (4.6, 24.5) 0.003 3.6 (-21.3, 36.5) 0.80 18.6 (-11.0, 58.0) 0.24 SumDiNCH Linear association 3.0 (1.0, 5.1) 0.003 0.9 (-5.2, 7.4) 0.77 0.4 (-5.9, 7.2) 0.90 Q2 (ref=Q1) -2.5 (-10.5, 6.3) 0.57 -15.2 (-35.3, 11.3) 0.24 2.9 (-22.4, 36.4) 0.84 Q3 (ref=Q1) 4.7 (-4.1, 14.3) 0.30 -17.7 (-37.6, 8.5) 0.17 23.4 (-7.4, 64.5) 0.15 Q4 (ref=Q1) 9.7 (0.6, 19.7) 0.04 -7.4 (-29.6, 21.9) 0.58 8.6 (-18.3, 44.4) 0.57 SumDEHTP Linear association 3.7 (1.2, 6.2) 0.004 3.6 (-4.1, 12.0) 0.37 0.5 (-7.3, 8.8) 0.91 Q2 (ref=Q1) -2.0 (-12.2, 9.5) 0.72 4.5 (-26.5, 48.4) 0.81 2.1 (-28.9, 46.6) 0.91 Q3 (ref=Q1) 6.4 (-4.7, 18.8) 0.27 30.0 (-8.5, 84.6) 0.14 -19.1 (-43.6, 16.1) 0.25 Q4 (ref=Q1) 9.0 (-2.6, 22.0) 0.13 8.8 (-24.0, 55.7) 0.65 5.1 (-27.4, 52.0) 0.79 72 Table 8 (cont’d). % Change (95%CI) in % Change (95%CI) in % Change (95%CI) in Biomarker P P P SumEstrogens SumTestosterones Estrogen/Androgen ratio MEP Linear association 1.9 (-0.4, 4.3) 0.11 9.3 (1.8, 17.5) 0.01 -10.0 (-16.4, -3.0) 0.01 Q2 (ref=Q1) 5.8 (-2.8, 15.1) 0.19 29.0 (-1.0, 68.2) 0.06 -17.1 (-37, 9.2) 0.18 Q3 (ref=Q1) 13.7 (4.4, 23.8) 0.003 40.5 (7.4, 83.8) 0.01 -29.2 (-46.4, -6.5) 0.02 Q4 (ref=Q1) 2.9 (-5.6, 12.1) 0.52 34.7 (2.7, 76.7) 0.03 -30.3 (-47.4, -7.7) 0.01 SumDBP Linear association 4.6 (1.3, 8.1) 0.01 -3.2 (-12.8, 7.4) 0.54 0.7 (-9.6, 12.2) 0.90 Q2 (ref=Q1) 12.1 (3.1, 22.0) 0.01 4.6 (-20.0, 36.7) 0.74 -2.5 (-26.2, 28.7) 0.86 Q3 (ref=Q1) 12.4 (3.4, 22.2) 0.01 17.1 (-10.1, 52.6) 0.24 -13.9 (-34.6, 13.3) 0.28 Q4 (ref=Q1) 11.0 (1.9, 21.0) 0.02 6.5 (-18.8, 39.8) 0.65 -7.8 (-30.4, 22.3) 0.57 SumDiBP Linear association 2.7 (-0.1, 5.6) 0.06 -3.9 (-12.0, 4.8) 0.37 0.5 (-8.2, 10.1) 0.91 Q2 (ref=Q1) 11.1 (2.2, 20.8) 0.01 31.2 (0.9, 70.7) 0.04 -17.2 (-37.1, 9) 0.18 Q3 (ref=Q1) 7.6 (-1.2, 17.2) 0.09 -7.5 (-29.2, 20.9) 0.57 -2.9 (-26.5, 28.5) 0.84 Q4 (ref=Q1) 11.1 (2.0, 20.9) 0.02 -8.5 (-29.8, 19.4) 0.51 5.2 (-20.3, 38.9) 0.72 SumPCP Linear association 2.9 (-0.3, 6.1) 0.08 5.6 (-4.3, 16.5) 0.28 -10.1 (-18.8, -0.6) 0.04 Q2 (ref=Q1) 8.4 (-0.4, 18.1) 0.06 31.2 (0.4, 71.6) 0.05 -22.5 (-41.3, 2.3) 0.07 Q3 (ref=Q1) 7.4 (-1.4, 17.0) 0.10 21.6 (-7.1, 59.1) 0.15 -18.4 (-38.2, 7.7) 0.15 Q4 (ref=Q1) 6.6 (-2.3, 16.3) 0.15 24.6 (-5.2, 63.8) 0.11 -31.9 (-48.7, -9.6) 0.01 Linear mixed models evaluated overall associations of phthalates/phthalate alternatives with ln-transformed hormones controlling for age, race/ethnicity, education, parity, smoking during 1st trimester, pre-pregnancy body mass index, diet quality, season of conception, fetal sex, and gestational age at hormone assessment. Phthalates/phthalate alternatives are either included as ln-transformed continuous variables or variables categorized into quartiles of exposure with Q1 as the reference group. β-estimates and 95%CIs for associations of continuous phthalates/phthalate alternatives with hormones were back- transformed to represent a % change in hormones for every 2-fold increase in phthalate/alternative. Associations where P ≤ 0.05 are bolded. CI, confidence interval; Q1-4, quartiles 1-4; Ref, reference. Phthalate/alternative concentrations in ng/mL or nmol/mL and hormone concentrations in ng/mL. 73 3.5.5. Gestational age- and fetal sex-specific associations of phthalate/alternative biomarkers with hormones Some associations between phthalate/alternative biomarkers and urinary SumEstrogens did differ by gestational age at hormone assessment and/or fetal sex (Figure 8). Positive associations of SumDEHP, MCPP, SumPlastics, and SumDiNCH with SumEstrogens were observed in both sexes at 8-15 weeks gestation, but also at 25-33 or 32-40 weeks gestation (Figures 8A, D, F, G). Positive associations of MCNP with SumEstrogens were also consistent in both sexes and strongest at 32-40 weeks gestation (Figure 8C). Associations of SumDEHTP with SumEstrogens were positive in women carrying female fetuses and observed at all three gestational timepoints (Figure 8H), while associations of SumDiNP with SumEstrogens were positive in women carrying female fetuses and strongest at 8-15 weeks gestation (Figure 8B). Positive associations of MEP, SumDiBP, and SumPCP with SumEstrogens were only observed in women carrying female fetuses and were strongest at 8-15 weeks gestation, but also 25-33 or 32-40 weeks gestation (Figures 8I, K, L), while positive associations of SumDBP with SumEstrogens were not fetal sex-specific at 8-15 weeks gestation, but were only observed in women carrying females at 25-33 weeks gestation (Figure 8J). In these stratified analyses, urinary concentrations of MBzP were not associated with SumEstrogens at any gestational timepoint or by fetal sex (Figure 8E). 74 Figure 8. Gestational age- and fetal sex-specific associations of phthalate and phthalate alternative molar sums or metabolites with urinary SumEstrogens. Linear mixed models controlled for age, race/ethnicity, education, parity, smoking during 1st trimester, pre-pregnancy body mass index, diet quality, season of conception, fetal sex, and gestational age at hormone assessment. Models also included a three-way and all relevant two-way interactions between fetal sex, gestational age at hormone assessment, and chemicals. Data are presented as the percent change (filled circle) and 95% CI (solid lines) in urinary SumEstrogens with every 2-fold increase in urinary phthalate or phthalate alternative biomarker. CIs that do not cross the null are significant at #P<0.1, *P<0.05, and **P<0.01. CI, confidence interval. Some associations of phthalate (but not phthalate alternative) biomarkers with urinary SumTestosterones also differed by gestational age and fetal sex (Figure 9). Positive associations of MEP and SumPCP with SumTestosterones were strongest in women carrying female fetuses at 8-15 weeks gestation along with 25-33 or 32-40 weeks gestation (Figure 9I, L). Additionally, a positive association was observed between SumPlastics and SumEstrogens in women carrying females at 8-15 weeks gestation (Figure 9F), while MBzP was positively associated with SumEstrogens in women carrying 75 females at 25-33 weeks gestation (Figure 9E). However, SumDiBP was negatively associated with SumTestosterones in women carrying male fetuses, with strongest associations observed at 8-15 and 25-33 weeks gestation (Figure 9K). Maternal SumDEHP, SumDiNP, MCNP, MCPP, SumDiNCH, SumDEHTP, and SumDBP were not associated with SumTestosterones, regardless of gestational timepoint of hormone assessment or fetal sex (Figures 9B, C, D, G, H, J). Figure 9. Gestational age- and fetal sex-specific associations of phthalate and phthalate alternative molar sums or metabolites with urinary SumTestosterones. Linear mixed models controlled for age, race/ethnicity, education, parity, smoking during 1st trimester, pre-pregnancy body mass index, diet quality, season of conception, fetal sex, and gestational age at hormone assessment. Models also included a three-way and all relevant two-way interactions between fetal sex, gestational age at hormone assessment, and chemicals. Data are presented as the percent change (filled circle) and 95% CI (solid lines) in urinary SumTestosterones with every 2-fold increase in urinary phthalate or phthalate alternative biomarker. CIs that do not cross the null are significant at #P<0.1, *P<0.05, and **P<0.01. CI, confidence interval. 76 We also observed that some associations of phthalate (but not phthalate alternative) biomarkers with the Estrogen/Androgen ratio differed by gestational age and/or fetal sex (Figure 10). Negative associations of MBzP, MEP, and SumPCP with the Estrogen/Androgen ratio were strongest in women carrying female fetuses at 25-33 or 32-40 weeks gestation (Figure 10E, I, L). However, associations of SumDEHP with Estrogen/Androgen ratio were positive in women carrying male fetuses and strongest at 8-15 and 32-40 weeks gestation (Figure 10A), while associations of SumDiBP with Estrogen/Androgen ratio were also positive in women carrying male fetuses, but only at 8-15 weeks gestation (Figure 10K). In stratified analyses, associations of SumDiNP, MCNP, MCPP, SumPlastics, SumDiNCH, SumDEHTP, and SumDBP with Estrogen/Androgen ratio did not differ by gestational age or fetal sex (Figures 10B, C, D, F, G, H, J). 77 Figure 10. Gestational age- and fetal sex-specific associations of phthalate and phthalate alternative molar sums or metabolites with urinary Estrogen/Androgen ratio. Linear mixed models controlled for age, race/ethnicity, education, parity, smoking during the 1st trimester, pre-pregnancy body mass index, diet quality, season of conception, fetal sex, and gestational age at hormone assessment. Models also included a three-way and all relevant two-way interactions between fetal sex, gestational age at hormone assessment, and chemicals. Data are presented as the percent change (filled circle) and 95% CI (solid lines) in urinary Estrogen/Androgen ratio with every 2-fold increase in urinary phthalate or phthalate alternative biomarker. CIs that do not cross the null are significant at #P<0.1, *P<0.05, and **P<0.01. CI, confidence interval. 3.6. DISCUSSION 3.6.1. Summary of major findings Our study suggests that select phthalate concentrations in pregnancy are associated with higher maternal urinary SumEstrogens and SumTestosterones, and a lower Estrogen/Androgen ratio. Additionally, we found that two biomarkers of phthalate alternatives (SumDiNCH and SumDEHTP) were positively associated with SumEstrogens, but not with SumTestosterones or the Estrogen/Androgen ratio. Some associations of phthalate/alternative biomarkers with urinary hormones tended to be 78 linear, with the strongest relationships observed at higher quartiles of phthalate or phthalate alternative biomarker concentrations. Importantly, many associations of phthalate and phthalate alternative biomarkers with SumEstrogens tended to be strongest in early and mid-to-late gestation and in women carrying females, while gestational age- and fetal sex-specific associations of phthalate/alternative biomarkers with SumTestosterones and Estrogen/Androgen ratio were less consistent. These findings further confirm that phthalates may have endocrine disrupting properties in pregnant women, which may have important public health implications for maternal and child life- long health. Our findings also support the need for additional studies evaluating the potential endocrine disrupting capacity of newer phthalate alternatives. 3.6.2. Assessment of gestational estrogens and testosterones in urine Our study is one of few that measured gestational hormone metabolite concentrations in urine. Validation studies in pre-menopausal women found that urinary estrogen metabolite concentrations have high within-person reproducibility (167). Other studies in pre-menopausal women also suggest that associations of certain reproductive and lifestyle factors, including age at first birth and smoking status, with urinary estrogens are consistent with those assessing plasma or serum estrogens (157, 158). However, findings from other studies suggest that evaluating urinary hormones may require a different interpretation compared to those evaluating plasma or serum hormone concentrations. For example, in pre-menopausal women, breast cancer risk was not associated with luteal plasma estrogens, was positively associated with follicular plasma estrogens, but was negatively associated with urinary estrogens (161). Additionally, a nested case- 79 control study of pregnant women found that women with pre-eclampsia had higher urinary estradiol concentrations than controls (160), which is inconsistent with studies evaluating estradiol in serum. Authors hypothesized that urinary estrogens may be markers of hormone metabolism rather than direct measurements of circulating concentrations (160), which further supports the use of summative measures of urinary estrogens in our study rather than individual urinary metabolites. This is evident in studies assessing concentrations of estrone, estradiol, and estriol in blood or urine showing that estradiol and estriol are the parent estrogens with the highest concentrations in blood and urine, respectively (4, 164). Urine and blood likely capture different hormone forms where unconjugated hormones are measured in plasma or serum, while conjugated hormones are generally measured in urine (161). Whether this is also the case in pregnant women (where the placenta is the major source of steroid hormones) will need to be confirmed in additional studies. However, urine may allow researchers to measure different types of hormone metabolites that cannot be measured in plasma or serum. Compared to blood sampling, urine may also provide opportunities for more extensive cross-pregnancy assessment of hormonal disruption in response to environmental exposures. Despite this, while our study findings suggest that phthalates may disrupt maternal hormones, some caution may be warranted when directly comparing the directionality/magnitude of our findings to studies where hormones were measured in maternal circulation. 3.6.3. Phthalate/alternatives are endocrine disruptors that alter gestational hormone levels Our findings related to urinary SumEstrogens are consistent with in vitro studies showing 80 that phthalates are weakly estrogenic (19, 20), which is concerning since higher maternal third trimester circulating estrone and estradiol concentrations have been associated with higher risk of breast cancer in mothers years after pregnancy (168). Our findings that some phthalates are associated with higher urinary SumTestosterones may be concerning given that elevated second and third trimester maternal testosterone levels may be associated with higher risk of pre-eclampsia and gestational diabetes (8, 9). However, these findings in pregnant women are inconsistent with in vivo studies reporting anti-androgenic effects of maternal DEHP and DBP exposure in male offspring (21, 22) or reduced late pregnancy blood testosterone levels in pregnant dams with DEHP exposure (169). These inconsistencies may be because these studies evaluated testosterone levels in male offspring rather than mothers or because they exposed pregnant dams to phthalates at doses irrelevant to humans (i.e. 100 mg/kg/day). It is also possible that urinary testosterone levels are not directly comparable to plasma testosterone concentrations, thus our findings should be corroborated using repeated plasma sampling. The Estrogen/Androgen ratio may be a relevant indicator of placental P450 aromatase activity, which is required to convert testosterone to estradiol (170), and our findings evaluating this ratio in urine may align with experimental models demonstrating that some phthalates reduce aromatase activity during gestation (171). Most importantly, our study is one of the first to show that biomarkers of phthalate alternatives (SumDiNCH and SumDEHTP) may exert similar endocrine disrupting effects on gestational hormones as those of the phthalate parent compounds they replace (e.g. DEHP), which has not been shown in previous studies (152, 172). Given the importance of estrogens and testosterone for pregnancy outcomes, substantially more needs to be 81 understood about the impacts of phthalates and their replacements on these hormones in pregnant women and the consequences of these disruptions for maternal and child health. 3.6.4. Associations of phthalate/alternative biomarkers with maternal hormones differed by gestational age The current study appears to be the first to evaluate associations of phthalate/alternative biomarkers with urinary markers of early, middle, and late gestation estrogen and testosterone concentrations. Associations of phthalate biomarkers with urinary estrogens at the three evaluated timepoints suggest that these exposures may be targeting placental hormonal pathways (4), which is consistent with experimental studies showing that phthalates modulate placental estrogen receptor activity and gene expression (173). However, our results are somewhat inconsistent with those from other prospective pregnancy cohort studies (77, 150-152, 155), which may be due to differences in assessing sex steroid concentrations in urine versus blood. One study of U.S. pregnant women found that urinary phthalate metabolites were positively associated with early pregnancy serum estrone or estradiol (consistent with our findings), but negatively associated with early pregnancy serum free testosterone (inconsistent with our findings) (150). However, a recent study in Michigan pregnant women found that urinary phthalate metabolites were not associated with first trimester plasma estrone, estradiol, estriol, or testosterone, but observed a negative association between MBP and maternal plasma estrone measured before delivery (155). Additionally, in a cohort of Puerto Rican pregnant women, phthalate/alternative biomarkers were not associated with maternal second 82 trimester serum estradiol or estriol, while MHBP and MEP were positively and negatively, respectively, associated with mid-pregnancy testosterone (77, 152). This same study also found that some associations of phthalate/alternative biomarkers with pregnancy serum estriol and testosterone differed by gestational age, such that positive and negative associations were observed in early and late second trimester, respectively (152). However, given that estrogens and testosterones have specific patterns of increases across all pregnancy trimesters, our study design may better represent hormone disruption at three key gestational timepoints. Additionally, unlike the previous studies described here, we assessed phthalate metabolites in a pooled urine sample across pregnancy, which may more accurately represent average gestational exposure to these chemicals. Nevertheless, to corroborate our findings, additional studies are needed that simultaneously assess (and compare) urinary and blood hormone concentrations at multiple key timepoints in pregnancy that correspond to important developmental windows. 3.6.5. Associations of phthalate/alternative biomarkers with maternal hormones differed by fetal sex Associations of prenatal phthalate metabolite concentrations with hormonally-driven pregnancy and birth outcomes, including pre-eclampsia, gestational diabetes, birth weight, and pre-term birth, often differ by fetal sex (174). However, prior to our study, only two studies in U.S. pregnant women evaluated fetal sex-specific associations of phthalates with gestational hormones, and their findings were inconsistent with ours (151, 155). In the multi-center cohort of U.S. women, SumDEHP was negatively associated 83 with mid-to-late estradiol in women carrying females (151). Additionally, this study found that SumDEHP, MBzP, and MBP were negatively associated with mid-to-late free/total testosterone in women carrying females, but MEP was positively associated with free/total testosterone in women carrying males (151). However, a study of Michigan pregnant women found that associations of urinary phthalate metabolite concentrations with plasma estrone, estradiol, estriol, or testosterone measured during the first trimester or before delivery were not fetal sex-specific (155). In addition to differences in the biological medium used for hormone assessment, these inconsistencies may be related to exposure measurement, as these studies quantified phthalate metabolites from individual spot urine samples collected during pregnancy. Additionally, these other studies may have been underpowered (n=180 for the multicenter U.S. cohort and n=121 for the Michigan cohort) to detect sex-specific associations of various phthalate biomarkers with hormones (151, 155). Consequently, future studies are needed to corroborate fetal sex-specific findings from both previous studies as well as from the current study. 3.6.6. Strengths and limitations First, using a pooled sample to assess pregnancy exposure to phthalates/alternatives means that there is uncertainty regarding the directionality of associations as some hormone measures were obtained prior to some exposure measures. However, using a pooled sample of five first morning urines for quantification of nonpersistent chemicals reduces exposure measurement error, provides a more stable measure of mean gestational exposure, and may, in fact, be a better reflection of exposure at any given 84 timepoint during pregnancy (162, 175). Second, while we did not validate urinary gestational sex-steroid hormones with those measured in serum or plasma, other studies in non-pregnant populations have established urine as a reliable biomarker for sex-steroid hormone assessment (156). Third, although urinary hormones and their metabolites may not directly reflect hormone levels in circulation, we only evaluated associations of phthalate/alternative biomarkers with the sum of estrogen and testosterone metabolites to characterize maternal hormonal disruption (161). Fourth, our findings may not be generalizable to more diverse population since our midwestern U.S. population of pregnant women is predominately white and highly educated. However, this rather homogenous population allows us to easily evaluate and propose biological pathways that can later be confirmed in more diverse populations and in appropriate experimental models. Fifth, while we accounted for urine dilution (i.e. hydration status) by specific gravity adjusting urinary analyte concentrations, specific gravity can vary by physiologic factors such body composition (176). However, specific gravity has low within-person variability, which makes it the more favorable marker of hydration status in pregnant populations relative to creatinine or osmolality (176). Sixth, while there may be residual or unmeasured confounding unaccounted for in our statistical models, we used a priori consideration and previous literature to make informed decisions about covariate selection. Seventh, there may be increased type I error because we did not adjust for multiple comparisons. However, our focus was on a qualitative interpretation of the findings, especially the stratified results, to identify trends in associations of phthalates/alternatives with hormones by gestational age at hormone assessment and fetal sex that will guide future research (166). Lastly, our study was limited to estrogens 85 and testosterones based on previous experimental studies, but studies in pregnant women suggest that phthalates can also alter maternal thyroid hormone, progesterone, and corticotropin-releasing hormone levels (77, 152). Given that phthalates may impact multiple hormonal pathways, future studies may be needed that consider a larger array of gestational hormones. 3.7. CONCLUSIONS To our knowledge, this is the first study to evaluate associations of phthalate and phthalate alternative biomarkers with cross-pregnancy gestational sex-steroid hormones measured in urine. Our study suggests that phthalates and their replacements may have endocrine disrupting capacity during pregnancy, and that some of these associations differ by gestational age and fetal sex. Interestingly, our findings combined with results from previously published experimental and observational studies suggest that the direction of these associations remains inconsistent, likely because of numerous factors related to study design and data analysis. Given that pregnant women are exposed to multiple phthalates/alternatives, future studies should evaluate the combined or interactive effects of multiple phthalates/alternatives on gestational hormone concentrations. Most importantly, because altered gestational sex-steroid hormone levels are associated with numerous adverse pregnancy and fetal outcomes, pregnant women may benefit from limiting their use of phthalate and phthalate replacement-containing products during pregnancy. 86 CHAPTER 4: ASSOCIATIONS OF INDIVIDUAL AND CUMULATIVE URINARY PHTHALATE AND REPLACEMENT BIOMARKERS WITH GESTATIONAL WEIGHT GAIN THROUGH LATE PREGNANCY This article/chapter has been published in Science of the Total Environment; Volume 855; Pacyga DC, Patti M, Papandonatos GD, Haggerty DK, Braun JM, Gardiner JC, Calafat AM, Schantz SL, Strakovsky RS; Associations of individual and cumulative phthalate and replacement biomarkers with gestational weight gain through late pregnancy; Copyright Elsevier (2022); https://doi.org/10.1016/j.scitotenv.2022.158788. 4.1. ABSTRACT Phthalates and their replacements are endocrine/metabolic disruptors that may impact gestational weight gain (GWG) - a pregnancy health indicator. We investigated overall and fetal sex-specific associations of individual and cumulative phthalate/replacement biomarkers with GWG. Illinois women (n = 299) self-reported their weight pre-pregnancy and at their final obstetric appointment before delivery (median 38 weeks). We calculated pre-pregnancy body mass index and gestational age-specific GWG z-scores (GWGz). We quantified 19 phthalate/replacement metabolites (representing 10 parent compounds) in pools of up-to-five first-morning urine samples, collected approximately monthly between 8 and 40 weeks gestation. We used linear regression, quantile-based g- computation (QGComp), and weighted quantile sum regression (WQSR) to evaluate associations of ten biomarkers (individual metabolites or parent molar-sums) individually or as mixtures (in interquartile range intervals) with GWGz. We evaluated associations in all women and stratified by fetal sex. Individually, sums of metabolites of di(2-ethylhexyl) 87 phthalate (ƩDEHP), di(isononyl) cyclohexane-1,2-dicarboxylate (ƩDiNCH), and di(2- ethylhexyl) terephthalate (ƩDEHTP) had consistent inverse associations with GWGz, and some associations were fetal sex-specific. When evaluating phthalates/replacements as a mixture, QGComp identified ƩDEHP, ƩDEHTP, and mono-(3-carboxypropyl) phthalate, along with sum of di(isononyl) phthalate metabolites (ƩDiNP) and monobenzyl phthalate as notable contributors to lower and higher GWGz, respectively, resulting in a marginal inverse joint association in all women (β: -0.29; 95% CI: -0.70, 0.12). In women carrying females, ƩDEHP contributed to the marginal inverse joint association (β: -0.54; 95% CI: - 1.09, 0.03). However, there was no overall association in women carrying males (β: 0.00; 95% CI: -0.60, 0.59), which was explained by approximately equal negative (driven by ƩDEHTP) and positive (driven by ƩDiNP) partial associations. WQSR analyses consistently replicated these QGComp findings. Biomarkers of phthalates/replacements were fetal sex-specifically associated with GWGz. Because ƩDEHTP contributed substantively to mixture associations, additional studies in pregnant women may be needed around this plasticizer replacement. 4.2. KEYWORDS DEHTP; DiNCH; fetal sex; gestational weight gain; phthalates; pregnancy. 4.3. INTRODUCTION Gestational weight gain (GWG) is an important, easily monitored indicator of maternal and fetal health. Deviations from the Institute of Medicine (IOM) GWG guidelines, including both excessive and inadequate GWG, may negatively impact health (3). For 88 example, compared to women with adequate GWG, those with inadequate GWG are at higher risk of delivering pre-term or small-for-gestational age newborns, whereas women with excessive GWG are at higher risk of developing gestational hypertension and having large-for-gestational age newborns (40). Long-term risks of inappropriate GWG include increased likelihood of maternal postpartum depression (43), excessive offspring adiposity (44), and greater maternal postpartum weight retention, all of which could lead to increased risk of later life-associated health complications (177, 178). Therefore, it is critical to identify modifiable risk factors associated with inappropriate GWG. Phthalates are a class of chemicals found in a wide variety of consumer products, including (but not limited to) food packaging materials, coatings of medications and supplements, personal care products, and cosmetics, with ubiquitous exposure among pregnant women in the U.S. (14). This is particularly concerning given that certain phthalates have been implicated in pregnancy-related metabolic disorders, including gestational hypertension (25, 179, 180) and diabetes (181-183). Industry use of plasticizer replacements such as cyclohexane-1,2-dicarboxylic acid diisononyl ester (DiNCH, a non-phthalate alternative) and di(2-ethylhexyl) terephthalate (DEHTP) appear to be on the rise (65, 184, 185). These replacements may have similar hormonally- mediated adverse health impacts to the original phthalates (83, 132, 186-188). Pregnancy and fetal development, including GWG, are regulated by coordinated hormonal, inflammatory, and metabolic processes, which may be biological targets of phthalates and their replacements (189-191). 89 Current evidence evaluating associations of prenatal urinary concentrations of phthalate metabolites (used as biomarkers of phthalate exposure) with GWG is inconclusive (31, 32, 192-196). For example, studies from the U.S. and China observed positive associations of first, second, or third trimester monoethyl phthalate (MEP) concentrations with excessive total GWG (31, 32, 195) and first trimester GWG (193). However, other studies reported that elevated second trimester urinary levels of low molecular weight phthalate biomarkers (driven by MEP) were associated with inadequate total GWG (192). Additionally, a study from China reported lower first or second trimester MEP concentrations among women with inadequate compared to those with adequate total GWG (194). It has become increasingly important to understand the cumulative impacts of phthalates/replacements and the role of each individual chemical within the context of the others. Of the few studies evaluating cumulative associations of phthalate biomarker mixtures with GWG, one study from China reported that a hazard index estimating exposures to di(2-ethylhexyl) phthalate (DEHP), diethyl phthalate (DEP), and di-n-butyl phthalate (DBP) across all pregnancy trimesters was associated with higher odds of excessive total GWG with most prominent associations observed for first trimester phthalate biomarkers (31). However, a Boston study observed no associations between sums of first, second, and third trimester phthalate metabolites of parents used in personal care products (MEP and monobutyl phthalate (MBP)), or phthalate metabolites shown to have anti-androgenic activity (MBP, monobenzyl phthalate (MBzP), mono- isobutyl phthalate (MiBP), mono(2-ethyl-5-hydroxyhexyl) phthalate (MEHHP), mono(2- 90 ethyl-5-carboxypentyl) phthalate (MECPP), mono(2-ethyl-5-oxohexyl) phthalate (MEOHP), and mono(2-ethylhexyl) phthalate (MEHP)) with total GWG (32). Finally, using Bayesian Kernel Machine Regression (BKMR), which can estimate the health effects of a complex mixture composed of highly correlated chemicals (197), a study of U.S. women observed that a mixture of select first trimester phthalate, bisphenol, and paraben biomarker concentrations was positively associated with total GWG, and this association was primarily driven by the molar sum of four DEHP metabolites (196). Additional studies are needed to better understand the individual and cumulative associations of phthalate metabolite biomarkers with GWG. Furthermore, previous studies only evaluated a limited number of phthalates. Studies assessing a larger panel of phthalates to which pregnant women are exposed, as well as common plasticizer replacements (such as DiNCH and DEHTP), are needed. Therefore, our overall objective was to evaluate associations of phthalate/replacement biomarkers with GWG through late pregnancy by considering phthalate biomarkers individually and as a mixture. We hypothesized that individually and as a mixture, phthalates/replacements would be associated with higher GWG. GWG is a complex phenotype with both maternal and fetal contributions (198), and fetal sex appears to be an important determinant of GWG (41, 199-201). Therefore, for our secondary objective we hypothesized that individually and as a mixture, associations of phthalates/replacements would differ by fetal sex. 91 4.4. MATERIALS AND METHODS 4.4.1. Illinois Kids Development Study (I-KIDS) recruitment and enrollment The current study includes pregnant women recruited from two local obstetric clinics in Champaign-Urbana, Illinois who were invited to participate in I-KIDS – a prospective pregnancy and birth cohort with the overarching aim of evaluating the impacts of prenatal chemical exposures on infant neurodevelopment. Recruitment and enrollment have been described in detail elsewhere (86). I-KIDS includes women who were≤ 15 weeks pregnant at enrollment, 18-40 years old, fluent in English, in a low-risk and singleton pregnancy, living within a 30-minute drive of the University of Illinois campus, and not planning to move out of the area before their child’s first birthday. The current study includes a sample of 303 women who enrolled in I-KIDS between February 2015 and August 2018, remained in the study through the birth of their infant, and have measurable concentrations of at least one urinary phthalate/replacement metabolite and pre- and late- pregnancy weights to calculate GWG through late pregnancy. As we have previously described, women in the current analytic sample are representative of the full I-KIDS cohort (202). All women provided written informed consent, and the study was approved by the Institutional Review Board at the University of Illinois. The analysis of de-identified specimens at the Centers for Disease Control and Prevention (CDC) laboratory was determined not to constitute engagement in human subjects’ research. 4.4.2. Assessment of urinary phthalate/replacement biomarker concentrations Women collected at least three and up to five first-morning urine samples in polypropylene urine cups at 8-15, 13-22, 19-28, 25-33, 32-40 weeks gestation (median 92 13, 17, 23, 28, and 34 weeks gestation, respectively), which corresponded to study home visits or routine prenatal care visits. Most of the 303 women contributed all five urine samples (94%), whereas 6% contributed three or four urine samples. Specifically, 99% of women provided a urine sample at 8-15 weeks gestation, 100% at 13-22 weeks gestation, 98% at 19-28 weeks gestation, 98% at 25-33 weeks gestation, and 98% at 32- 40 weeks gestation. We have previously described urine collection, processing, and storage protocols in detail (86, 202). For the current study, we quantified phthalate/replacement metabolite concentrations and measured specific gravity from pooled samples of up to five individual first-morning urines collected from each woman. Pooled samples were shipped overnight to the CDC Division of Laboratory Sciences in two batches in chronological order of participant enrollment as follows: enrolled February 2015 - July 2016 and enrolled July 2016 - August 2018. Using previously published methods with rigorous quality assurance/quality control protocols and excellent long-term reproducibility (104-106), the following 19 phthalate/replacement metabolites were quantified: monocarboxynonyl phthalate (MCNP), monocarboxyoctyl phthalate (MCOP), monooxononyl phthalate (MONP), monoisononyl phthalate (MiNP), MEHP, MEHHP, MEOHP, MECPP, mono-(3-carboxypropyl) phthalate (MCPP), MBzP, MBP, mono- hydroxybutyl phthalate (MHBP), MiBP, mono-hydroxy-isobutyl phthalate (MHiBP), MEP, cyclohexane-1,2-dicarboxylic acid-mono(carboxyoctyl) ester (MCOCH), cyclohexane- 1,2-dicarboxylic acid-monohydroxy isononyl ester (MHiNCH), mono(2-ethyl-5- hydroxyhexyl) terephthalate (MEHHTP), and mono(2-ethyl-5-carboxypentyl) terephthalate (MECPTP). 93 4.4.3. Collection and calculation of GWG z-scores through late pregnancy Within 24 hours of delivery, women reported their measured weight (in pounds) and the date of their most recent obstetric appointment (median: 38, range: 28 – 41 weeks gestation). We subtracted pre-pregnancy weight (which was self-reported during a home visit at 8 – 15 weeks gestation) from the weight reported at the last obstetric appointment before delivery to calculate GWG through late pregnancy (in kg). In the few cases where weight at the last obstetric visit prior to delivery was not available (n = 17), we used the reported weight from an earlier obstetric visit (median: 34, range: 29 – 37 weeks gestation). We calculated pre-pregnancy BMI- and gestational age-specific GWG z- scores based on previously validated methods using a reference population of pregnant women from Europe, North America, and Oceania (203). 4.4.4. Collection of maternal sociodemographic, lifestyle, and health information After enrollment (8 – 15 weeks gestation), study staff conducted home visits to interview women about their race/ethnicity, educational attainment, annual household income, parity, smoking in the first trimester, alcohol intake in the first trimester, as well as pre- pregnancy weight and height. Self-reported pre-pregnancy weight and height were used to calculate pre-pregnancy BMI (kg/m2). To ascertain early pregnancy stress status, women completed the Perceived Stress Scale, a 10-item questionnaire that asks about thoughts and feelings during the past month (204, 205). At 8 – 15 and 32 – 40 weeks gestation, women also completed a semi-quantitative food frequency questionnaire (FFQ) that was adapted from the full length Block-98 FFQ (NutritionQuest, Berkeley, CA) and validated in a pregnant population (101, 206, 207). Reported dietary intakes 94 representing diet during the previous three months were used to calculate the mean of early and mid-to-late pregnancy Alternative Healthy Eating Index (AHEI-2010). The AHEI- 2010 is an 11-component diet quality index (out of 110 total points) based on foods and nutrients shown to be predictive of chronic disease risk and mortality, where a higher score indicates better overall diet quality (102, 103). 4.4.5. Statistical analysis Derivation of the analytic sample is presented in Figure 11. Out of 303 women with data on concentrations of all phthalate/replacement biomarkers and GWG z-scores, 299 (148 and 151 carrying males and females, respectively) were included in final covariate- adjusted single- and multi-pollutant analyses. Characteristics of the sample are presented as n (%) or median (range). Figure 11. Derivation of analytic sample for evaluating associations of phthalate/replacement biomarkers with GWG z-scores. The chart presents sample sizes for sensitivity analyses and analyses evaluating unadjusted and adjusted associations of phthalate/replacement biomarkers with GWG z-scores using multiple linear regression models. Sample sizes for QGComp and WQSR models are those reported for adjusted models. GWG, gestational weight gain; QGComp, quantile-based g-computation; WQSR, weighted quantile sum regression. We evaluated specific gravity-adjusted phthalate/replacement biomarkers as molar sums or individual metabolites, which reflect exposures to 10 phthalate/replacement parent 95 compounds (86, 202). We created phthalate/replacement molar sums (in nmol/mL) by summing metabolites from common precursors as follows: MEHP, MEHHP, MEOHP, and MECPP for the sum of DEHP metabolites (ƩDEHP); MCOP, MiNP, and MONP for the sum of metabolites of di-isononyl phthalate (ƩDiNP); MBP and MHBP for the sum of DBP metabolites (ƩDBP); MiBP and MHiBP for the sum of di-iso-butyl phthalate metabolites (ƩDiBP); MHiNCH and MCOCH for the sum of DiNCH metabolites (ƩDiNCH); and MEHHTP and MECPTP for the sum of DEHTP metabolites (ƩDEHTP). Specifics about equations are published elsewhere (86). Molar concentrations were back-converted to ng/mL by multiplying ƩDEHP, ƩDiNP, ƩDBP, ƩDiBP, ƩDiNCH, and ƩDEHTP by the molecular weights of MECPP, MCOP, MBP, MiBP, MHiNCH, and MECPTP, respectively (113, 186, 202). We estimated exposure to di-isodecyl phthalate, di-n-octyl phthalate, benzylbutyl phthalate (BBzP), and DEP using ng/mL concentrations of their corresponding urinary metabolites MCNP, MCPP, MBzP, and MEP, respectively. Based on the previous literature, we considered an extensive number of potential covariates in statistical models evaluating associations of phthalate/replacement biomarkers with GWG z-scores (31, 32, 192-196). We generated a directed acyclic graph (DAG, Figure 12) (208), which included covariates that we and others found to be associated with both phthalate/replacement biomarkers and GWG z-scores. We used the DAG to guide the minimum sufficient adjustment set of covariates (208). We assessed correlations between covariates to test for potential multicollinearity, but all covariates were only weakly or moderately correlated (r < 0.4; data not shown). Therefore, all final covariate-adjusted single-pollutant and mixtures models accounted for race/ethnicity, 96 educational attainment, annual household income, smoking in the first trimester, pre- pregnancy BMI, and maternal diet quality. Annual household income, pre-pregnancy BMI, and maternal diet quality were included as continuous variables, whereas race/ethnicity, educational attainment, and smoking in the first trimester were categorized with the reference groups indicated in Table 9. We specified models that additionally accounted for perceived stress, parity, and alcohol intake in the first trimester, but our results were relatively unchanged with the inclusion of these variables (data not shown). 97 Figure 12. Directed acyclic graph for associations of phthalate/replacement biomarkers with GWG z-scores. Phthalate/replacement biomarkers are the exposures (green/black circle) and gestational weight gain is the outcome (blue/black circle). Green circles indicate variables associated with the exposure, blue circles indicate variables associated with the outcome, and red circles indicate variables associated with both the exposure and outcome. Gray circles represent latent variables, while white circles represent variables that were included in final covariate-adjusted analyses (minimum sufficient adjustment set). BMI, body mass index; GWG, gestational weight gain. 98 4.4.6. Primary and sensitivity single-pollutant analyses To address our main objectives, we first specified multivariable linear regression models to evaluate single-pollutant associations of phthalate/replacement biomarkers with GWG z-scores. All phthalate/replacement biomarkers were ln-transformed because of their right-skewed distributions. For concentrations below the limit of detection (LOD) we used instrumental reading values to avoid bias associated with imputing values $60,000 (73%). Most women did not smoke (95%) or consume alcohol (58%) in the first trimester, around half had overweight or obesity before pregnancy (48%), and median 104 AHEI-2010 was 55.9 (min, max: 28.1, 80.2). Around two thirds of women reported having low perceived stress, whereas the rest reported having moderate or high perceived stress in the first trimester. Median GWG through median 38 weeks gestation was 15.0 kg (min, max: -10.9, 38.1). 105 Table 9. Characteristics of the study sample (n=303). Category n (%) or median (min - max) Age, years 31.0 (18.0 - 40.0) 1 Race/ethnicity Non-Hispanic white (ref) 247 (81.5) 2Others 56 (18.5) 1 Educational attainment Some college or less 52 (17.2) College grad or higher (ref) 251 (82.8) 1 Annual household income < $60,000 82 (27.2) $60,000 - $99,999 113 (37.5) ≥ $100,000 106 (35.2) Parity 0 children 161 (53.2) 1 child 94 (31.0) ≥ 2 children 48 (15.8) 1 Smoking in the first trimester No (ref) 288 (95.1) Yes 15 (4.9) Alcohol intake in the first trimester None 176 (58.1) Any alcohol consumed 127 (41.9) 1 Pre-pregnancy BMI Under-/normal weight (< 25 kg/m2) 158 (52.2) Overweight (25 - 29.9 kg/m2) 78 (25.7) Obese (≥ 30 kg/m2) 67 (22.1) Perceived stress in the first trimester Low stress (0-13 pts) 188 (62.7) Moderate stress (14-26 pts) 106 (35.3) High stress (27-40 pts) 6 (2.0) Fetal sex Female 153 (50.5) Male 150 (49.5) Birth weight (grams) Females 3360.8 (2353.0, 4507.6) Males 3611.7 (2588.3, 4394.2) AHEI-2010 55.9 (28.1 - 80.2) GWG through late pregnancy (kg) 15.0 (-10.9 - 38.1) GWG through late pregnancy (z-scores) 0.5 (-3.0 - 4.1) Gestational age at pre-delivery weight report (weeks) 38.9 (28.9 - 41.9) 1Included in final statistical models as covariates. 2Includes Hispanic, Black, Asian, American Indian, Multiracial, Other. Percentages may not add up to 100% due to missing. Subset sample (n missing): annual household income, GWG z-scores, gestational age at late pregnancy weight (2 missing); perceived stress (3 missing). AHEI-2010, Alternative Healthy Eating Index 2010 scored out of 110; GWG, gestational weight gain. 106 4.5.2. Urinary phthalate/replacement metabolite biomarker concentrations in I- KIDS The distribution of urinary phthalate/replacement biomarker concentrations are presented in Table 10. Greater than 96% of women had detectable concentrations of at least one phthalate metabolite per parent compound (including DEHTP), and > 90% of women had detectable concentrations of at least one DiNCH metabolite. As we have shown previously, most phthalate/replacement biomarkers were weakly-to-moderately correlated with each other (r < 0.4), though we did observe a strong correlation between ƩDiNP and MCPP (r > 0.8) (202). As reported previously, median uncorrected urinary phthalate/DiNCH biomarker concentrations in I-KIDS were similar to those of same age women from the National Health and Nutrition Examination Survey (NHANES) during a similar time period (202). However, DEHTP metabolite concentrations in I-KIDS were 1.5 to 3 times higher than those from NHANES (202). 107 Table 10. Distribution of maternal urinary specific gravity-adjusted phthalate/ replacement biomarker concentrations (n = 303). % 25th percentile 50th percentile 75th percentile Parent Biomarker detectable (ng/mL) (ng/mL) (ng/mL) Di-isodecyl phthalate, DiDP MCNP* 100.0 1.33 1.81 2.58 ƩDiNP* -- 4.94 7.96 15.11 MCOP 100.0 4.47 7.09 13.65 Di-isononyl phthalate, DiNP MiNP 31.4 0.4 0.62 1.09 MONP 100.0 1.71 2.64 4.63 ƩDEHP* -- 14.75 19.71 29.89 MEHP 75.6 0.85 1.32 2.13 Di(2-ethylhexyl) phthalate, MEHHP 100.0 3.62 5.40 8.32 DEHP MEOHP 100.0 2.88 4.00 6.34 MECPP 100.0 6.08 8.32 12.64 Di-n-octyl phthalate, DOP MCPP* 96.0 0.86 1.30 1.90 Benzylbutyl phthalate, BBzP MBzP* 99.3 2.54 5.16 10.00 ƩDBP* -- 9.36 14.05 19.22 Di-n-butyl phthalate, DBP MBP 100.0 8.46 12.68 17.41 MHBP 90.1 0.75 1.22 1.87 ƩDiBP* -- 7.86 11.56 19.59 Di-iso-butyl phthalate, DiBP MiBP 99.7 5.74 8.53 14.40 MHiBP 99.7 2.13 3.11 5.47 Diethyl phthalate, DEP MEP* 100.0 14.12 27.47 47.79 ƩDiNCH* -- 1.16 1.88 3.43 Diisononyl-cyclohexane-1,2- MHiNCH 90.7 0.69 1.14 2.26 dicarboxylate, DiNCH MCOCH 67.3 0.48 0.71 1.22 ƩDEHTP* -- 30.18 69.65 158.54 Di(2-ethylhexyl) MEHHTP 100.0 3.84 8.16 20.11 terephthalate, DEHTP MECPTP 100.0 25.69 58.88 135.53 All concentrations are presented in ng/mL. All metabolite concentrations were specific gravity-adjusted. Molar sums were converted back to ng/mL by multiplying each molar sum by the molecular weight of a corresponding major metabolite as discussed in the statistical analysis section. *Indicates the biomarkers of interest for the current study. MBP, monobutyl phthalate; MBzP, monobenzyl phthalate; MCNP, monocarboxynonyl phthalate; MCOCH, cyclohexane-1,2-dicarboxylic acid-mono(carboxyoctyl) ester; MCOP, monocarboxyoctyl phthalate; MCPP, mono- (3-carboxypropyl) phthalate; MECPP, mono(2-ethyl-5-carboxypentyl) phthalate; MECPTP, mono(2-ethyl-5- carboxypentyl) terephthalate; MEHP, mono(2-ethylhexyl) phthalate; MEHHP, mono(2-ethyl-5-hydroxyhexyl) phthalate; MEHHTP, mono(2-ethyl-5-hydroxyhexyl) terephthalate; MEOHP, mono(2-ethyl-5-oxohexyl) phthalate; MEP, monoethyl phthalate; MHBP, mono-hydroxybutyl phthalate; MHiBP, mono-hydroxy-isobutyl phthalate; MHiNCH, cyclohexane-1,2-dicarboxylic acid-monohydroxy isononyl ester; MiBP, mono-isobutyl phthalate; MiNP, monoisononyl phthalate; MONP, monooxononyl phthalate; ƩDiNP, sum of di(isononyl) phthalate metabolites; ƩDEHP, sum of di(2-ethylhexyl) phthalate metabolites; ƩDBP, sum of di-n-butyl phthalate metabolites; ƩDiBP, sum of di-iso-butyl phthalate metabolites; ƩDiNCH, sum of di(isononyl) cyclohexane-1,2-dicarboxylate metabolites; ƩDEHTP, sum of di(2-ethylhexyl) terephthalate metabolites. 4.5.3. Associations of individual phthalate/replacement biomarkers with GWG z- scores In single-pollutant linear regression models, ƩDEHP and ƩDEHTP were inversely associated with GWG z-scores (Table 11). For example, IQR increases in ƩDEHP and 108 ƩDEHTP were associated with -0.16 (95% CI: -0.29, -0.03) and -0.20 (95% CI: -0.38, - 0.02) lower GWG z-scores, respectively. MCPP and ƩDiNCH were also meaningfully associated with GWG z-scores, where IQR increases in MCPP and ƩDiNCH were associated with -0.09 (95% CI: -0.21, 0.03) and -0.11 (95% CI: -0.22, 0.01) lower GWG z-scores, respectively. These results persisted in sensitivity analyses excluding the four women with GWG < 0 (Supplemental Table 12). We also observed suggestive differences in associations between some phthalate/replacement biomarkers and GWG z-scores by pre-pregnancy BMI (Supplemental Table 13). However, when additionally evaluating differences by pre-pregnancy BMI and fetal sex, no definitive patterns emerged and results were notably less precise (Supplemental Table 14). Table 11. Single-pollutant associations of phthalate/replacement biomarkers with GWG z-scores. All women Women carrying males Women carrying females (n = 299) (n = 148) (n = 151) Biomarker β (95% CI) β (95% CI) β (95% CI) Pbiomarker*sex MCNP 0.00 (-0.12, 0.13) 0.03 (-0.16, 0.23) 0.00 (-0.16, 0.17) 0.80 ƩDiNP 0.01 (-0.13, 0.15) 0.25 (0.02, 0.49) -0.08 (-0.25, 0.09) 0.02 ƩDEHP -0.16 (-0.29, -0.03) 0.04 (-0.14, 0.22) -0.35 (-0.53, -0.18) 0.002 MCPP -0.09 (-0.21, 0.03) -0.03 (-0.22, 0.17) -0.11 (-0.27, 0.05) 0.49 MBzP 0.09 (-0.06, 0.25) 0.15 (-0.08, 0.37) 0.03 (-0.18, 0.25) 0.46 ƩDBP -0.09 (-0.22, 0.05) -0.08 (-0.28, 0.11) -0.10 (-0.27, 0.07) 0.90 ƩDiBP -0.03 (-0.16, 0.10) 0.01 (-0.20, 0.23) -0.05 (-0.22, 0.12) 0.64 MEP -0.08 (-0.24, 0.07) 0.00 (-0.22, 0.22) -0.13 (-0.35, 0.09) 0.39 ƩDiNCH -0.11 (-0.22, 0.01) -0.17 (-0.35, 0.01) -0.04 (-0.19, 0.11) 0.27 ƩDEHTP -0.20 (-0.38, -0.02) -0.26 (-0.50, -0.02) -0.12 (-0.37, 0.13) 0.43 Data are presented as the change (95% CI) in GWG z-scores for every IQR increase in phthalate/replacement biomarker concentration. Models accounted for race/ethnicity, educational attainment, annual household income, smoking in the 1st trimester, pre-pregnancy BMI, and maternal diet quality. To obtain fetal sex-specific estimates and the interaction P-value (Pbiomarker*sex), we additionally included a multiplicative interaction between phthalate/replacement biomarker and fetal sex. Cl, confidence interval; GWG, gestational weight gain; MCNP, monocarboxynonyl phthalate; ƩDiNP, sum of di(isononyl) phthalate metabolites; ƩDEHP, sum of di(2-ethylhexyl) phthalate metabolites; MCPP, mono-(3-carboxypropyl) phthalate; MBzP, monobenzyl phthalate; ƩDBP, sum of di- n-butyl phthalate metabolites; ƩDiBP, sum of di-iso-butyl phthalate metabolites; MEP, monoethyl phthalate; ƩDiNCH, sum of di(isononyl) cyclohexane-1,2-dicarboxylate metabolites; ƩDEHTP, sum of di(2-ethylhexyl) terephthalate metabolites. 109 Table 12. Sensitivity analyses for single-pollutant associations of phthalate/replacement biomarkers with GWG z-scores. Unadjusted associations in all Excluding women with women (n = 301) GWG < 0 (n = 295)1 Biomarker β (95% CI) β (95% CI) MCNP 0.00 (-0.12, 0.13) 0.03 (-0.09, 0.15) ƩDiNP 0.01 (-0.12, 0.15) 0.05 (-0.09, 0.19) ƩDEHP -0.20 (-0.32, -0.07) -0.14 (-0.27, -0.02) MCPP -0.08 (-0.20, 0.05) -0.07 (-0.19, 0.05) MBzP 0.04 (-0.11, 0.20) 0.08 (-0.07, 0.23) ƩDBP -0.11 (-0.24, 0.01) -0.10 (-0.22, 0.03) ƩDiBP -0.05 (-0.18, 0.08) -0.04 (-0.16, 0.09) MEP -0.14 (-0.29, 0.01) -0.03 (-0.18, 0.12) ƩDiNCH -0.11 (-0.22, 0.01) -0.09 (-0.20, 0.02) ƩDEHTP -0.22 (-0.39, -0.05) -0.17 (-0.34, 0.00) Data are presented as the change (95% CI) in GWG z-scores for every IQR increase in phthalate/replacement biomarker concentration. 1Accounted for race/ethnicity, educational attainment, annual household income, smoking in the 1st trimester, pre-pregnancy BMI, and maternal diet quality. BMI, body mass index; CI, confidence interval; GWG, gestational weight gain; MCNP, monocarboxynonyl phthalate; ƩDiNP, sum of di(isononyl) phthalate metabolites; ƩDEHP, sum of di(2-ethylhexyl) phthalate metabolites; MCPP, mono-(3-carboxypropyl) phthalate; MBzP, monobenzyl phthalate; ƩDBP, sum of di-n-butyl phthalate metabolites; ƩDiBP, sum of di-iso-butyl phthalate metabolites; MEP, monoethyl phthalate; ƩDiNCH, sum of di(isononyl) cyclohexane-1,2- dicarboxylate metabolites; ƩDEHTP, sum of di(2-ethylhexyl) terephthalate metabolites. 110 Table 13. Pre-pregnancy BMI-specific associations of individual phthalate/replacement biomarkers with GWG z- scores. All women Obese Overweight Under-/normal weight (n = 299) (n = 67) (n = 77) (n = 155) Biomarker β (95% CI) β (95% CI) β (95% CI) β (95% CI) Pint MCNP 0.00 (-0.12, 0.13) -0.20 (-0.50, 0.10) 0.17 (-0.10, 0.44) 0.00 (-0.16, 0.16) 0.20 ƩDiNP 0.01 (-0.13, 0.15) -0.10 (-0.36, 0.16) 0.13 (-0.17, 0.43) 0.02 (-0.18, 0.22) 0.50 ƩDEHP -0.16 (-0.29, -0.03) -0.30 (-0.55, -0.05) -0.03 (-0.28, 0.22) -0.16 (-0.35, 0.03) 0.31 MCPP -0.09 (-0.21, 0.03) -0.14 (-0.41, 0.12) -0.04 (-0.31, 0.23) -0.09 (-0.26, 0.07) 0.86 MBzP 0.09 (-0.06, 0.25) 0.03 (-0.31, 0.37) 0.28 (-0.01, 0.57) 0.02 (-0.20, 0.24) 0.33 ƩDBP -0.09 (-0.22, 0.05) 0.11 (-0.19, 0.41) -0.15 (-0.40, 0.10) -0.12 (-0.29, 0.06) 0.37 ƩDiBP -0.03 (-0.16, 0.10) 0.06 (-0.15, 0.27) -0.04 (-0.32, 0.24) -0.13 (-0.34, 0.08) 0.45 MEP -0.08 (-0.24, 0.07) -0.13 (-0.44, 0.17) -0.22 (-0.54, 0.11) 0.00 (-0.22, 0.22) 0.52 ƩDiNCH -0.11 (-0.22, 0.01) -0.09 (-0.29, 0.10) -0.24 (-0.47, 0.00) -0.04 (-0.22, 0.14) 0.42 ƩDEHTP -0.20 (-0.38, -0.02) -0.29 (-0.66, 0.08) -0.36 (-0.70, -0.03) -0.08 (-0.33, 0.17) 0.36 Data are presented as the change (95% CI) in GWG z-scores for every IQR increase in phthalate/replacement biomarker concentration in all women and by the following pre-pregnancy BMI categories: BMI ≥ 30 kg/m2 (obese), BMI 25-29.9 kg/m2 (overweight), BMI < 25 kg/m2 (under-/normal weight). All models accounted for race/ethnicity, educational attainment, annual household income, smoking in the 1st trimester, pre-pregnancy BMI, and maternal diet quality. Pre-pregnancy BMI-specific findings, including the interaction P-value (Pint), were obtained from models that also included a multiplicative interaction between biomarker and pre-pregnancy BMI. BMI, body mass index; CI, confidence interval; GWG, gestational weight gain; MCNP, monocarboxynonyl phthalate; ƩDiNP, sum of di(isononyl) phthalate metabolites; ƩDEHP, sum of di(2-ethylhexyl) phthalate metabolites; MCPP, mono-(3-carboxypropyl) phthalate; MBzP, monobenzyl phthalate; ƩDBP, sum of di-n-butyl phthalate metabolites; ƩDiBP, sum of di-iso-butyl phthalate metabolites; MEP, monoethyl phthalate; ƩDiNCH, sum of di(isononyl) cyclohexane-1,2-dicarboxylate metabolites; ƩDEHTP, sum of di(2-ethylhexyl) terephthalate metabolites. 111 Table 14. Pre-pregnancy BMI- and fetal sex-specific associations of individual phthalate/replacement biomarkers with GWG z-scores. Women carrying males Under-/normal weight All (n = 148) Obese (n = 38) Overweight (n = 39) (n = 71) Biomarker β (95% CI) β (95% CI) β (95% CI) β (95% CI) MCNP 0.03 (-0.16, 0.23) -0.28 (-0.79, 0.23) 0.10 (-0.32, 0.53) 0.09 (-0.16, 0.34) ƩDiNP 0.25 (0.02, 0.49) 0.07 (-0.32, 0.46) 0.28 (-0.16, 0.71) 0.42 (0.00, 0.84) ƩDEHP 0.04 (-0.14, 0.22) -0.14 (-0.43, 0.15) 0.08 (-0.24, 0.39) 0.22 (-0.11, 0.55) MCPP -0.03 (-0.22, 0.17) 0.11 (-0.28, 0.49) -0.09 (-0.42, 0.25) -0.07 (-0.38, 0.25) MBzP 0.15 (-0.08, 0.37) -0.01 (-0.46, 0.45) 0.22 (-0.17, 0.61) 0.20 (-0.13, 0.52) ƩDBP -0.08 (-0.28, 0.11) 0.31 (-0.09, 0.71) -0.36 (-0.68, -0.04) -0.06 (-0.36, 0.25) ƩDiBP 0.01 (-0.20, 0.23) 0.08 (-0.32, 0.47) 0.09 (-0.50, 0.67) -0.03 (-0.31, 0.26) MEP 0.00 (-0.22, 0.22) 0.11 (-0.31, 0.52) -0.13 (-0.62, 0.36) 0.00 (-0.31, 0.32) ƩDiNCH -0.17 (-0.35, 0.01) -0.10 (-0.44, 0.25) -0.23 (-0.55, 0.10) -0.19 (-0.46, 0.09) ƩDEHTP -0.26 (-0.50, -0.02) -0.10 (-0.58, 0.38) -0.29 (-0.75, 0.18) -0.36 (-0.71, -0.02) Women carrying females Under-/normal weight All (n = 151) Obese (n = 29) Overweight (n = 38) (n = 84) Biomarker β (95% CI) β (95% CI) β (95% CI) β (95% CI) MCNP 0.00 (-0.16, 0.17) -0.12 (-0.50, 0.26) 0.27 (-0.09, 0.63) -0.06 (-0.27, 0.16) ƩDiNP -0.08 (-0.25, 0.09) -0.19 (-0.54, 0.15) 0.05 (-0.35, 0.46) -0.08 (-0.31, 0.15) ƩDEHP -0.35 (-0.53, -0.18) -0.79 (-1.25, -0.33) -0.16 (-0.54, 0.22) -0.31 (-0.53, -0.10) MCPP -0.11 (-0.27, 0.05) -0.33 (-0.70, 0.05) 0.09 (-0.37, 0.55) -0.10 (-0.29, 0.10) MBzP 0.03 (-0.18, 0.25) 0.09 (-0.42, 0.61) 0.44 (0.01, 0.88) -0.15 (-0.45, 0.14) ƩDBP -0.10 (-0.27, 0.07) -0.21 (-0.68, 0.25) 0.19 (-0.21, 0.59) -0.15 (-0.35, 0.06) ƩDiBP -0.05 (-0.22, 0.12) 0.10 (-0.16, 0.36) 0.00 (-0.34, 0.34) -0.29 (-0.60, 0.02) MEP -0.13 (-0.35, 0.09) -0.38 (-0.81, 0.05) -0.18 (-0.64, 0.28) 0.02 (-0.29, 0.32) ƩDiNCH -0.04 (-0.19, 0.11) -0.06 (-0.31, 0.19) -0.22 (-0.55, 0.12) 0.07 (-0.16, 0.30) ƩDEHTP -0.12 (-0.37, 0.13) -0.53 (-1.10, 0.04) -0.45 (-0.92, 0.03) 0.22 (-0.12, 0.57) Data are presented as the change (95% CI) in GWG z-scores for every IQR increase in phthalate/replacement biomarker concentration in all women and by pre-pregnancy BMI [BMI ≥ 30 kg/m2 (obese), BMI 25-29.9 kg/m2 (overweight), BMI < 25 kg/m2 (under-/normal weight)] and fetal sex [males, females]. All models accounted for race/ethnicity, educational attainment, annual household income, smoking in the 1st trimester, pre-pregnancy BMI, and maternal diet quality. Pre-pregnancy BMI-specific findings, including the interaction P-value (Pint), were obtained from models that also included a multiplicative three-way interaction (and all relevant two-way interactions) between biomarker, pre-pregnancy BMI, and fetal sex. BMI, body mass index; CI, confidence interval; GWG, gestational weight gain; MCNP, monocarboxynonyl phthalate; ƩDiNP, sum of di(isononyl) phthalate metabolites; ƩDEHP, sum of di(2-ethylhexyl) phthalate metabolites; MCPP, mono-(3-carboxypropyl) phthalate; MBzP, monobenzyl phthalate; ƩDBP, sum of di-n-butyl phthalate metabolites; ƩDiBP, sum of di-iso-butyl phthalate metabolites; MEP, monoethyl phthalate; ƩDiNCH, sum of di(isononyl) cyclohexane-1,2-dicarboxylate metabolites; ƩDEHTP, sum of di(2- ethylhexyl) terephthalate metabolites. 112 4.5.4. Associations of phthalate/replacement biomarker mixtures with GWG z- scores When using QGComp to evaluate the joint association of the phthalate/replacement biomarker mixture with GWG z-scores, we observed that simultaneous IQR increases in all biomarker concentrations in the mixture were marginally associated with lower GWG z-scores (β: -0.29; 95% CI: -0.70, 0.12) (Figure 13 and Table 15), as the partial negative (β: -1.02) and positive (β: 0.73) associations were similar in strength. The largest phthalate/replacement biomarker contributors to partial negative associations were ƩDEHP (28%), ƩDEHTP (23%), and MCPP (18%), whereas the largest biomarker contributors to partial positive associations were ƩDiNP (64%) and MBzP (36%) (Table 16). 113 Figure 13. Joint associations of phthalate/replacement biomarkers with GWG z-scores in a) all women, b) women carrying males, and c) women carrying females. QGComp models and WQSR models using partial reverse scoring approach accounted for race/ethnicity, educational attainment, annual household income, smoking in the 1st trimester, pre- pregnancy BMI, and maternal diet quality. Data are presented as the change (square for QGComp, circle for WQSR models) and 95% CI (vertical black lines) in GWG z-scores for an IQR change in the mixture. Separate models were specified for all women (n = 299), women carrying males (n = 148), and women carrying females (n = 151). QGCompneg, scaled effect size in the negative direction; QGComppos, scaled effect size in the positive direction; QGCompneg+pos, joint association as the sum of QGCompneg and QGComppos. WQSneg and WQSpos, sum of negative and positive weights, respectively, calculated after renormalizing positive and negative weights to 1.0; WQSneg+pos, joint association as the sum of WQSneg and WQSpos. GWG, gestational weight gain; QGComp, quantile g-computation; WQSR, weighted quantile sum regression. 114 Table 15. Cumulative associations of phthalate/replacement mixture with GWG z-scores from QGComp and WQSR models. All women Women carrying males Women carrying females (n = 299) (n = 148) (n = 151) QGComp1 β (95% CI) or β β (95% CI) or β β (95% CI) or β QGCompneg -1.02 -1.14 -1.07 QGComppos 0.73 1.14 0.54 QGCompneg+pos -0.29 (-0.70, 0.12) 0.00 (-0.60, 0.59) -0.54 (-1.09, 0.03) WQSR Reverse Scoring Method2 β (95% CI) or β β (95% CI) or β β (95% CI) or β WQSneg -0.56 (-1.00, -0.14) -0.63 (-1.02, -0.23) -0.44 (-1.00, 0.12) WQSpos 0.35 (0.08, 0.62) 0.69 (0.25, 1.12) 0.26 (-0.07, 0.58) WQSneg+pos -0.21 0.06 -0.19 WQSR Standard Method3 β (95% CI) β (95% CI) β (95% CI) WQSneg -0.30 (-0.65, 0.05) -0.20 (-0.75, 0.30) -0.50 (-0.90, -0.05) WQSpos -0.02 (-0.35, 0.35) 0.35 (-0.05, 0.80) -0.30 (-0.75, 0.15) Data are presented as the change (95% CI) in GWG z-scores for an IQR change in the mixture. Separate models were specified for all women (n = 299), women carrying males (n=148), and women carrying females (n=151). 1QGComp models. QGCompneg, scaled effect size in the negative direction; QGComppos, scaled effect size in the positive direction; QGCompneg+pos, cumulative associations as the sum of QGCompneg and QGComppos. 2Negatively constrained WQSR models using the reverse scoring method where chemicals that were positively associated with GWG z-scores were reverse coded. WQSneg and WQSpos, sum of negative and positive weights, respectively, calculated after renormalizing positive and negative weights to 1.0; WQSneg+pos, sum of WQSneg and WQSpos. 3Standard WQSR models where separate negatively (WQSneg) and positively (WQSpos) constrained models were evaluated. GWG, gestational weight gain; QGComp, quantile g-computation; WQSR, weighted quantile sum regression. 115 Table 16. Relative weights from reverse scoring WQSR and QGComp models evaluating associations of the phthalate/replacement mixture with GWG z-scores. All women Women carrying males Women carrying females (n = 299) (n = 148) (n = 151) Negative Positive Negative Positive Negative Positive Method Chemical (wt) Chemical (wt) Chemical (wt) Chemical (wt) Chemical (wt) Chemical (wt) ƩDEHP (0.284) ƩDiNP (0.639) ƩDBP (0.304) ƩDiNP (0.518) ƩDEHP (0.553) ƩDiNP (0.372) ƩDEHTP (0.233) MBzP (0.361) ƩDEHTP (0.292) MBzP (0.386) ƩDEHTP (0.122) MBzP (0.355) MCPP (0.175) MCPP (0.151) ƩDiBP (0.096) MEP (0.118) ƩDBP (0.273) MCNP (0.117) MCNP (0.097) MCNP (0.076) QGComp1 MEP (0.094) ƩDiNCH (0.087) ƩDiBP (0.072) ƩDiNCH (0.046) MEP (0.036) ƩDiNCH (0.038) ƩDiBP (0.032) ƩDEHP (0.034) MCPP (0.022) ƩDBP (0.020) ƩDEHP (0.250) ƩDiNP (0.500) ƩDBP (0.383) ƩDiNP (0.358) ƩDEHP (0.372) ƩDBP (0.342) ƩDEHTP (0.190) MBzP (0.427) ƩDEHTP (0.255) MBzP (0.249) MEP (0.177) MBzP (0.265) WQSR MCPP (0.175) MCNP (0.073) MCPP (0.203) ƩDiBP (0.136) ƩDiBP (0.153) ƩDEHTP (0.199) Reverse MEP (0.128) ƩDiNCH (0.159) MCNP (0.094) ƩDiNCH (0.139) MCNP (0.194) Scoring ƩDiNCH (0.113) MEP (0.092) MCPP (0.084) Method2 ƩDiBP (0.075) ƩDEHP (0.073) ƩDiNP (0.074) ƩDBP (0.070) Separate models were specified for all women (n = 299), women carrying males (n=148), and women carrying females (n=151). Bolded chemicals are those that crossed the threshold (1/10 chemicals in the mixture). 1Data are presented as the chemical (relative weight) from QGComp models. 2Data are presented as the chemical (relative weight) from negatively constrained WQSR models using the reverse scoring method where chemicals that were positively associated with GWG z-scores (see positive column) were reverse coded; weights were renormalized to 1.0 within positive and negative columns. GWG, gestational weight gain; MCNP, monocarboxynonyl phthalate; ƩDiNP, sum of di(isononyl) phthalate metabolites; ƩDEHP, sum of di(2-ethylhexyl) phthalate metabolites; MCPP, mono-(3-carboxypropyl) phthalate; MBzP, monobenzyl phthalate; ƩDBP, sum of di-n-butyl phthalate metabolites; ƩDiBP, sum of di-iso-butyl phthalate metabolites; MEP, monoethyl phthalate; ƩDiNCH, sum of di(isononyl) cyclohexane-1,2-dicarboxylate metabolites; ƩDEHTP, sum of di(2-ethylhexyl) terephthalate metabolites; QGComp, quantile g-computation; WQSR, weighted quantile sum regression; Wt, weight. 116 Results from associations of all 10 phthalate/replacements as a cumulative mixture with GWG z-scores in all women using standard WQSR models are presented in Supplemental Table 15. When evaluating mixture effects via WQSR using the (partial) reverse scoring approach, so that all biomarkers are associated with the outcome in the same direction, we observed that the full mixture was associated with GWG z-scores (β: -0.90; 95% CI: -1.60, -0.20). Decomposing this overall association into contributions from originally negative and positive mixture components, we observed that IQR increases in the negative phthalate/replacement biomarkers were associated with 0.56 (95% CI: 0.14, 1.00) reductions in GWG z-scores (Figure 13a), with the largest contributors being ƩDEHP (25%), ƩDEHTP (19%), and MCPP (18%) (Table 16). Correspondingly, IQR increases in positive biomarkers were associated with 0.35 (95% CI: 0.08, 0.62) increases in GWG z-scores (Figure 13a), with the largest contributors being ƩDiNP (50%) and MBzP (43%) (Table 16). 4.5.5. Fetal sex-specific associations of phthalate/replacement biomarkers with GWG z-scores When evaluating associations using individual phthalate/replacement biomarkers, the inverse associations of ƩDiNCH and ƩDEHTP with GWG z-scores observed in all women appeared to be more robust in women carrying males, whereas inverse associations in women carrying females were modest (Table 11). However, inverse associations of ƩDEHP with GWG z-scores in all women were driven by women carrying females (β: - 0.35, 95% CI: -0.53, -0.18), with no association observed in those carrying males (β: 0.04, 95% CI: -0.14, 0.22). Additionally, we identified a positive association between ƩDiNP 117 and GWG z-scores in women carrying males (β: 0.25, 95% CI: 0.02, 0.49), with no association observed in those carrying females (β: -0.08, 95% CI: -0.25, 0.09). Using QGComp, we observed that the phthalate/replacement biomarker mixture was marginally associated with lower GWG z-scores in women carrying females (β: -0.54; 95% CI; -1.09, 0.03) (Figure 13c), but not associated with GWG z-scores in women carrying males (β: 0.00; 95% CI: -0.60, 0.59) (Figure 13b). In women carrying females, ƩDEHP (55%) was identified as the largest biomarker contributor to the partial negative association (Table 16). However, in women carrying males, the magnitude of the partial negative (β: -1.14) and partial positive (β: 1.14) associations were equal in strength, resulting in no overall joint association of the mixture with GWG z-scores in this sub- sample (Figure 13b and Table 15). In women carrying males, ƩDBP (30%) and ƩDEHTP (29%) were the largest contributors to partial negative associations, while ƩDiNP (52%) and MBzP (39%) were the largest contributors to partial positive associations (Table 16). Results evaluating associations of all 10 phthalate/replacements as a cumulative mixture with GWG z-scores separately in women carrying males and females using standard WQSR models are presented in Table 15. When evaluating associations of the phthalate/replacement biomarker mixture with GWG z-scores using WQSR with the reverse scoring method, we observed that the biomarker mixture was associated with GWG z-scores both in women carrying males and females, but the association was much more robust in women carrying males (β: -1.30; 95% CI: -2.15, -0.50) than in those carrying females (β: -0.70; 95% CI: -1.55, 0.20) (Figure 13b,c and Table 15). With 118 regards to the partial negative associations stratified by fetal sex, we observed that IQR increases in the negative biomarkers were associated with GWG z-scores decreases of 0.63 (95% CI: 0.23, 1.02) and 0.44 (95% CI: -0.12, 1.00) in women carrying males and females, respectively. In women carrying males, the strongest phthalate/replacement biomarker contributors to associations in the negative direction were ƩDBP (39%) and ƩDEHTP (26%), whereas in women carrying females, the strongest contributors were ƩDEHP (37%) and MEP (18%) (Table 16). When evaluating partial positive associations by fetal sex, we observed that IQR increases in positive biomarkers were associated with 0.69 (95% CI: 0.25, 1.12) higher GWG z-scores in women carrying males (Figure 13b), with ƩDiNP (36%) and MBzP (25%) identified as the strongest contributors to this association (Table 16). In women carrying females, IQR increases in positive biomarkers were weakly associated with 0.26 (95% CI: -0.07, 0.58) higher GWG z-scores (Figure 13c), with ƩDBP (34%) was identified as the strongest contributor to this association (Table 16). 4.6. DISCUSSION We observed that urinary biomarker concentrations of phthalate plasticizers and their replacements were associated with GWG z-scores, which was generally consistent between single-pollutant and mixture modeling approaches, as well as between QGComp and WQSR. In secondary single- and multi-pollutant analyses, we observed that some associations of phthalates/replacement biomarkers with GWG z-scores were fetal-sex- specific, a finding which has not been reported previously. In mixtures models, while consistent marginal inverse associations were observed in women carrying females, in 119 those carrying males, we observed equal associations in both the negative and positive directions. These results contribute to the growing evidence that exposure to phthalates may alter GWG in pregnant women and provide novel information about the potential for widely used plasticizer replacements, DiNCH and DEHTP, to impact GWG. 4.6.1. Individual phthalate/replacement biomarkers are associated with GWG z- scores Contrary to our hypothesis, in our study, select phthalate biomarkers (ƩDEHP and MCPP) and biomarkers of replacements (ƩDiNCH and ƩDEHTP) were generally inversely associated with GWG z-scores, and for some of these biomarkers, associations were either more prominent in women carrying females (for ƩDEHP) or males (for ƩDiNCH and ƩDEHTP). Interestingly, a positive association emerged for ƩDiNP with GWG z-scores in women carrying males. It is important to view GWG as a complex phenotype, with contributions from the fetus, placenta, amniotic fluid, maternal fluid (i.e., blood, extracellular fluid), protein and fat storage, as well as uterine and breast tissues (198). Our findings in women carrying females support the idea that some phthalates may target biological pathways that prevent weight gain in one or more of these storage depots (3). In pregnancy, exposures to certain phthalates are associated with altered metabolic processes required for appropriate GWG, such as glucose and lipid homeostasis (212, 213). Specifically, in experimental models, phthalates have been shown to interact with peroxisome proliferator-activated receptor gamma, liver X receptors, and retinoid X receptors, which are important regulators of glucose and lipid homeostasis (23). These metabolic processes are regulated by sex-steroid hormones (such as estrogen and 120 progesterone) and cytokines/adipokines produced by the maternal-fetal-placental unit (3), which are also potential biological targets of phthalates/replacements (86, 173, 174, 214- 217). For example, phthalates have been shown to interfere with the regulation of the hypothalamic-pituitary-gonadal (HPG) axis by interacting with steroidogenic enzymes and by modulating the activity of hormone receptors (e.g. estrogen receptor, androgen receptor) (218). The hormone-mediated role of the placenta in GWG and the known sex differences in the placental response to phthalates may also explain our fetal sex-specific findings (173, 214), although further studies are needed to elucidate the biological basis for these observations. Nevertheless, appropriate GWG is critical for maternal health and fetal development (3), and both insufficient and excessive GWG may have adverse implications for maternal and offspring health. For example, women with insufficient GWG may be more likely to experience postpartum depression (43), whereas those with excessive GWG are at higher risk of postpartum weight retention that could lead to cardiometabolic disease later in life (2). Babies of mothers with insufficient GWG have higher odds of being born pre- term and small-for-gestational age, whereas those of mothers with excessive total GWG have higher odds of macrosomia or being born large-for-gestational age (219). Importantly, these early birth phenotypes have been linked with the later development of respiratory problems, cognitive and/or behavioral deficits, and cardiometabolic disease in children (220-223). GWG modeled continuously has also been shown to be associated with these same adverse childhood outcomes. Specifically, one study showed that compared to a GWG z-score of 0.0, a GWG z-score of +0.5 was associated with 2.2 (95% 121 CI: 0.7, 3.7) excess cases of childhood overweight or obesity per 100 pregnancies (224). Another study found that compared to GWG z-scores between -1.0 and +1.0, children of women who had a GWG z-score over +1.0 spent 15.0 seconds (95% CI: 1.8, 28.0) longer completing a task measuring executive performance, suggesting that higher GWG is associated with poorer executive function performance in children (225). Phthalates are also sex-specifically associated with many of these same short- and long-term infant and child health outcomes (174), so it is possible that associations of certain phthalates with GWG could partially explain relationships of phthalates with child and maternal health, and this should be explored further. The current literature evaluating associations of individual phthalates with GWG is mixed, but also limited with regards to phthalate replacements and evaluating differences by fetal sex. Consistent with our findings, one study from the Netherlands evaluated GWG through late pregnancy and reported that associations of mid-pregnancy MCPP with GWG trended in the negative direction (192). However, most other studies evaluated total GWG using the IOM categories or trimester-specific GWG and quantified phthalate metabolite biomarkers from individual spot urines collected in early, mid, and/or late pregnancy, which limits direct comparisons to our study. Of the studies evaluating total GWG, two (from Anhui Province in China and Salinas Valley, California in the U.S.) observed that higher urinary metabolite concentrations of DEHP, DBP, and DEP in early/mid pregnancy were associated with higher total GWG and higher odds of excessive GWG (31, 195). Additionally, in pregnant women from Boston, mean urinary MEP concentrations were associated with higher odds of excessive total GWG, although this 122 relation was non-monotonic (32). Conversely, in the previously discussed Netherlands cohort, higher early/mid gestation urinary low molecular weight phthalate biomarkers were associated with higher odds of insufficient total GWG (192), while in women from Hubei Province in China, MEP concentrations were lower among women with inadequate compared to adequate total GWG (33). Inconsistent findings among observational studies could also be explained by to covariates accounted for in statistical models and study population characteristics. For example, both our study and the Netherlands study adjusted for maternal diet, but this was not accounted for by other studies. Additionally, the majority of our women are non-Hispanic white, of higher socioeconomic status (SES), and almost half had overweight or obesity before pregnancy, which may also explain differences in findings from cohorts in China (where the majority of women were normal weight before pregnancy) and California (where most women were migrants from Mexico with lower SES). The experimental evidence related to maternal body weight gain is also mixed. For example, rodent studies evaluating the effects of prenatal DEP exposure observed decreasing, increasing, and no effects on maternal body weight in response to DEP (38). Another study observed that compared to controls, F0 generation dams exposed to DEHP (pre-conception to weaning), DiNP (mating to weaning, pre-conception to weaning), or DBP (pre-conception to weaning) gained more weight (39). Further research can assist in identifying what may be contributing to these conflicting findings within and between observational and experimental studies. 123 4.6.2. Phthalate/replacement biomarkers as a cumulative mixture are associated with GWG z-scores Given that pregnant women are likely exposed to numerous phthalates/replacements, evaluating these chemicals as a cumulative mixture is important for understanding the aggregate association of many phthalates/replacement with pregnancy and fetal health (66). For the current study, we compared results from two widely used statistical mixtures methods, QGComp and WQSR (72, 210). To make our findings more comparable to those from QGComp, while simultaneously satisfying the WQSR assumption that all mixture components are associated with the outcome in the same direction, we utilized a (partial) reverse scoring approach for WQSR. Although QGComp and WQSR have distinct purposes and properties, we generally observed very consistent findings across the two methods – most notably, there was almost perfect agreement regarding the ordering and direction of the contribution of key phthalate/replacement biomarkers to joint mixture associations. Specifically, in all women, we observed that the partial negative and positive associations were similar in magnitude, with the overall biomarker mixture only marginally inversely associated with GWG z-scores through late pregnancy. ƩDEHP, ƩDEHTP, and MCPP were identified as the largest contributors to inverse associations and ƩDiNP and MBzP identified as the most prominent contributors to positive associations. When stratifying by fetal sex, we observed marginal inverse associations in women carrying females, a result that was largely driven by ƩDEHP. However, in women carrying males, we observed equal partial positive (driven by ƩDiNP) and negative (driven by ƩDEHTP) associations, which resulted in a negated joint association between the mixture and GWG z-scores in this sub-sample. Our results suggest that women carrying 124 males – more so than those carrying females – may be equally sensitive to chemicals that have opposing effects on GWG. Phthalates/replacements mainly exert their effects on the endocrine system by binding to hormone receptors, such as estrogen receptors alpha and beta that can be found and expressed in different quantities in multiple cell types, which could partially explain different and potentially opposing responses to endocrine disruption (226, 227). However, additional studies are needed to corroborate these fetal sex-specific mixture findings, as well as to understand what could explain these opposing chemical effects at the physiological level, especially in women carrying males. Our mixture results in all women are somewhat consistent with those of the Netherlands cohort observing modest inverse associations of urinary phthalic acid (a proxy for total phthalate exposure) with GWG through late pregnancy (192). Conversely, using a different method to estimate exposure, the cohort from Anhui Province calculated a hazard index by summing estimated intakes of DBP, BBzP, and DEHP and observed the cumulative index was positively associated with total GWG and odds of excessive total GWG (31). Additionally, using BKMR, a study of pregnant women attending a fertility clinic in Boston identified that first trimester DEHP metabolites, MiBP, and propyl paraben contributed most to positive associations between a cumulative mixture of multiple non- persistent chemical classes and total GWG, with DEHP metabolites being the largest contributors (196). Lastly, the sum of phthalate metabolite biomarkers categorized as anti- androgenic was not associated with total GWG in the Boston women (32). 125 As discussed earlier, differences in study characteristics may explain discrepancies between our findings and those from other studies. However, each study, including ours, used different statistical approaches and different proxies of cumulative phthalate exposure or included additional classes of non-persistent chemicals in the mixture, which could also explain inconsistencies across studies. Each statistical mixtures method has its own unique limitations, but also strengths. While a limitation of WQSR is that it assumes that all chemicals in the mixture are associated with the outcome in the same direction, this method is quite appropriate for estimating the cumulative impact of chemicals from distinct exposure sources (i.e., phthalates/replacements) and identifying the most prominent chemical contributors to this association (210). In contrast, QGComp is best suited for chemicals from a common exposure source, as it assumes that all exposures are changing in the same direction, not independently; on the other hand, it is able to simultaneously consider chemicals that are associated with the outcome of interest in different directions (72). Overall, currently available statistical mixtures approaches have been developed to address a variety of questions that may arise in the field of environmental epidemiology. In the case of associations between phthalates/replacement and GWG, future studies can consider using the above- mentioned and other statistical mixtures methods, including BKMR, which is a robust approach when associations between the chemical mixture and outcome of interest are complex (i.e., interaction, non-linearity) (197). 4.6.3. Limitations and Strengths This study has some limitations, but also several strengths. First, we were unable to 126 calculate total GWG, which limits our ability to compare our results to previous studies that considered total GWG or IOM clinical cut-offs. However, we used a previously validated method to calculate GWG z-scores that are standardized by pre-pregnancy BMI and gestational age at late pregnancy weight (203), which provides a valid assessment of gestational age- and BMI-specific GWG compared to raw measures alone when total GWG is not available. Second, our choice of using an international GWG reference chart to calculate GWG z-scores may have influenced our observed findings. However, we validated our findings by calculating GWG z-scores using a reference chart developed from a sample of Pittsburgh pregnant women (228, 229), and observed very consistent associations of phthalate/replacement biomarkers with GWG z-scores regardless of the reference chart (data not shown). Third, there are some limitations to using a pooled urine sample to quantify phthalates/replacement metabolite concentrations. We were unable to consider differences by the timing of exposure, which has been demonstrated in other studies (31, 192, 196). Also, we may have lost some temporality for evaluating associations of phthalates/replacements with GWG since our first urine sample was collected toward the end of the first trimester. However, given the non-persistent nature of phthalates/replacements in the body and relatively high within person variability in biomarker concentrations across pregnancy (162), pools of up to five first morning urine samples improved the stability of our exposure measure and better approximated total pregnancy exposure to phthalates/replacements compared to a single first morning sample. Fourth, given that I-KIDS is still ongoing, it will take some time to obtain information about the number of women who developed preeclampsia and gestational diabetes, which could influence observed associations. However, we expect very few 127 cases of pre-eclampsia and gestational diabetes since most women enrolled in I-KIDS did not have major preexisting conditions. Fifth, an important strength is that we evaluated phthalates/replacement biomarkers individually and in a cumulative mixture using both WQSR and QGComp – the latter two allowed us to estimate joint associations of multiple phthalates/replacement biomarkers with GWG z-scores and identify the relative biomarker contributors to these associations. Sixth, while we evaluated many individual- level maternal sociodemographic, lifestyle, and health factors, our results are subject to residual confounding. For example, we did not collect information pertaining to physical activity or sleep quantity/quality before or during pregnancy, which may be important determinants of GWG (230) and urinary phthalate biomarker concentrations (231, 232). There may also be concerns related to co-pollutant confounding. However, we conducted sensitivity analyses limiting our mixture to only include chemicals that crossed the threshold or only include chemicals with positive or negative β-estimates in single pollutant models and observed that all cumulative mixture associations with GWG z- scores remained consistent to what we reported (data not shown). Seventh, studies evaluating GWG typically consider pre-pregnancy BMI as an effect modifier (211), which was not a primary goal of our study because one of our objectives was to focus on fetal sex-specific associations, and we were therefore underpowered to examine associations stratified by fetal sex and pre-pregnancy BMI. However, this is the first study (to our knowledge) to propose and show fetal sex as an important moderator that should be considered in future studies. Lastly, most I-KIDS participants are non-Hispanic White women of relatively high SES, which limits the generalizability of our findings to other populations. However, urinary concentrations of most phthalate and replacement 128 metabolites were similar to those of same age women in the U.S. at similar time periods indicating that exposure in I-KIDS women is consistent with exposure in U.S. women. 4.7. CONCLUSION In our relatively high-SES sample of U.S. pregnant women from the Midwest, a mixture of phthalates that included plasticizer replacements DiNCH and DEHTP was marginally associated with lower GWG through late pregnancy. However, our fetal sex-specific findings suggest that women carrying males may be more sensitive to phthalates that may be associated with GWG in opposing directions than those carrying females. Additionally, DEHTP was an important contributor to mixture associations with GWG, which, along with our studies showing that DEHTP is increasing in our sample (202) and is associated with maternal hormonal disruption (86), highlights a potential concern for regrettable chemical substitution. Therefore, further research related to this plasticizer replacement is needed in pregnant populations with particular consideration for fetal sex. Finally, experimental animal studies may help elucidate the biological mechanisms underlying the interaction of phthalates/replacement, fetal sex, and GWG. 129 CHAPTER 5: CONCLUSIONS AND FUTURE DIRECTIONS 5.1. OVERALL CONCLUSIONS In this dissertation, we evaluated maternal risk factors (Chapter 2), pregnancy hormonal targets (Chapter 3), and cardiometabolic consequences (Chapter 4) of exposure to phthalates and their replacements in pregnancy. In Chapter 2, we characterized phthalate/replacement exposure and identified several sociodemographic, lifestyle, and reproductive characteristics, as well as seasonal and time trends predictive of a mixture of phthalate/replacement concentrations. Specifically, we identified two distinct clusters of women: those with low phthalate (including ∑DEHTP) and those with high phthalate (including ∑DEHTP) biomarker concentrations. We also identified four components representative of phthalate/replacement biomarker concentrations from common exposure sources or those that track with certain behaviors. We identified age, marital status, annual household income, parity, pre-pregnancy BMI, caffeine intake, conception season, and enrollment year as important predictors of k- means clusters and at least one PC. Additionally, race/ethnicity, education, employment, pregnancy intention, smoking and consuming alcohol in the first trimester, and first trimester diet quality were identified as important determinants of at least one PC. Our findings contribute additional information about predictors of phthalate/replacement mixtures that may be important confounding factors in studies evaluating associations of phthalates/replacements with pregnancy-related health outcomes, including pregnancy hormonal targets and cardiometabolic consequences. Furthermore, our results may inform future perinatal health recommendations by providing insights into both non- 130 modifiable and modifiable characteristics of pregnant women who are most likely to be exposed to these chemicals. In Chapter 3, we observed that select phthalates and replacements individually were associated with altered maternal urinary estrogen concentrations, and these associations differed by fetal sex and by pregnancy timepoint. Specifically, we showed that select phthalate concentrations in pregnancy were associated with higher maternal urinary SumEstrogens and SumTestosterones, and a lower Estrogen/Androgen ratio. Additionally, two biomarkers of phthalate alternatives (SumDiNCH and SumDEHTP) were positively associated with SumEstrogens, but not with SumTestosterones or the Estrogen/Androgen ratio. Some associations of phthalate/alternative biomarkers with urinary hormones tended to be linear, with the strongest relationships observed at higher quartiles of phthalate or phthalate alternative biomarker concentrations. Importantly, many associations of phthalate and phthalate alternative biomarkers with SumEstrogens tended to be strongest in early and mid-to-late gestation and in women carrying females, while gestational age- and fetal sex-specific associations of phthalate/alternative biomarkers with SumTestosterones and Estrogen/Androgen ratio were less consistent. These findings further confirm that phthalates may have endocrine disrupting properties in pregnant women, which may have important public health implications for maternal and child life-long health. Our findings also support the need for additional studies evaluating the potential endocrine disrupting capacity of newer phthalate alternatives. 131 In Chapter 4, we reported that a mixture of phthalates that included plasticizer replacements DiNCH and DEHTP was marginally associated with lower GWG through late pregnancy, but these associations were also fetal sex-specific, a finding which has not been reported previously. Specifically, associations between urinary biomarker concentrations of phthalate plasticizers and their replacements and GWG z-scores were generally consistent between single-pollutant and mixture modeling approaches, as well as between QGComp and WQSR. In secondary single- and multi-pollutant analyses, we observed that some associations of phthalates/replacement biomarkers with GWG z- scores were fetal-sex-specific. In mixtures models, while consistent marginal inverse associations were observed in women carrying females, in those carrying males, we observed equal associations in both the negative and positive directions. Our fetal sex- specific findings suggest that women carrying males may be more sensitive to phthalates associated with GWG in opposing directions than those carrying females. These results contribute to the growing evidence that exposure to phthalates may alter GWG in pregnant women and provide novel information about the potential for widely used plasticizer replacements, DiNCH and DEHTP, to impact GWG. Given the results presented in this dissertation, several notable findings warrant future research, which will be discussed in the next section. 132 5.2. FUTURE DIRECTIONS 5.2.1. Determine if associations between phthalates/replacements and GWG are mediated by urinary hormones In Chapters 2 and 3, we demonstrated that maternal urinary phthalate/replacement biomarker concentrations were associated with urinary hormones and GWG z-scores. A natural next step for these studies would be to determine whether urinary hormones mediate associations of phthalates and GWG z-scores. Given that urinary hormone concentrations were assessed repeatedly across pregnancy, we would be able to utilize multiple mediator approaches under the causal framework to consider the joint mediation of urinary hormone concentrations at 8-15, 25-33, 32-40 weeks gestation (233). We could also simply evaluate mediation of association between phthalates/replacements and GWG z-scores by each hormone timepoint separately. However, we would first need to determine if urinary hormones are associated with GWG z-scores. Based on biological plausibility, we would hypothesize that hormones, particularly estrogens, would be directly associated with GWG, such that increases in estrogens would lead to higher GWG (3, 4). Unfortunately, we may not be able to justify conducting a formal mediation analysis given that we showed phthalates/replacements were associated with higher urinary estrogens, but lower GWG z-scores. We may not be able to directly interpret our findings assessing urinary hormones as we would if we measured hormones in blood (this is further discussed in the next section). This is supported by one study in pre-menopausal women finding breast cancer risk was not associated with luteal plasma estrogens, was positively associated with follicular plasma estrogens, but was negatively associated with urinary estrogens (161). Furthermore, a nested case-control study of pregnant women found that 133 women with pre-eclampsia had higher urinary estradiol concentrations than controls (160), which is inconsistent with studies evaluating estradiol in serum. Therefore, future studies in I-KIDS will need to determine not only how best to model the repeated urinary hormone measurements as mediators, but also carefully consider whether mediation is biologically plausible when assessing sex-steroid hormones in urine. 5.2.2. Corroborate associations of phthalates/replacements with urinary hormones using blood hormone levels As discussed in Chapter 3, a major strength of quantifying hormones in urine was to provide a novel approach for assessing hormones repeatedly across pregnancy, which is often a major limitations of prior studies. Collecting repeated urine samples may also be easier and less invasive than collecting multiple blood samples, which would also ease participant burden. However, the literature assessing whether urinary hormones directly reflect circulating hormones suggests that urinary hormones may be markers of hormone metabolism rather than of circulating hormone concentrations (160). Urine and blood likely capture different hormone forms, where unconjugated hormones are measured in plasma or serum, whereas conjugated hormones are generally measured in urine (161). Unfortunately, in pregnancy, quantifying hormones in urine has not been extensively validated against the gold standard (blood), particularly in studies collecting both blood and urine consistently across gestation. In I-KIDS, blood was not collected at the three urine collection timepoints for hormone assessment (8-15, 25-33, 32-40 weeks gestation). Pregnant participants did provide a blood and urine sample at 13-22 weeks gestation. A future study could examine correlations between hormone concentrations quantified from 134 the 13-22 week urine and blood aliquots. While this would not directly help address the potential limitations of the study in Chapter 3, this would be one way to internally validate how well urine hormone levels reflect blood hormone levels. This is not to discount the utility of assessing urinary hormones in pregnancy since urine may allow researchers to measure different types of potentially biologically active hormone metabolites that cannot be measured in plasma or serum (164, 234). Additionally, compared to blood sampling, urine may also provide opportunities for more extensive cross-pregnancy assessment of hormonal disruption in response to environmental exposures. However, it will be important to further characterize urinary hormones, especially to determine what they reflect from a biological standpoint, which will aid in interpreting findings utilizing this approach. Future pregnancy cohort studies could also consider incorporating the collection of multiple blood and urine samples into their protocols to help address this. 5.2.3. Evaluate associations of a phthalate/replacement mixture with hormones In Chapter 3, we only evaluated associations of individual phthalates/replacements with urinary hormones, but whether a mixture of these chemicals is also associated with hormones is also warranted. Assessing phthalates/replacements one at a time with health outcomes makes it challenging to know the true impact of real-life exposure to chemical mixtures (66). To better simulate real life exposures, evaluating cumulative or joint associations of phthalates/replacements with hormones will be necessary to further understand the endocrine disrupting potential of these chemicals. There are many statistical mixtures methods available to address this, but selecting the best mixtures method depends on the primary research question since each statistical mixtures method 135 has its own unique limitations, but also strengths. We used WQSR and QGComp to evaluate associations of a phthalate/replacement mixture with GWG z-scores in Chapter 4. While a limitation of WQSR is that it assumes that all chemicals in the mixture are associated with the outcome in the same direction, this method is quite appropriate for estimating the cumulative impact of chemicals from distinct exposure sources (i.e., phthalates/replacements) and identifying the most prominent chemical contributors to this association (Czarnota et al., 2015). In contrast, QGComp is best suited for chemicals from a common exposure source, as it assumes that all exposures are changing in the same direction, not independently; on the other hand, it is able to simultaneously consider chemicals that are associated with the outcome of interest in different directions (Keil et al., 2020). Another widely used approach that we have not discussed is BKMR, which is a robust approach when associations between the chemical mixture and outcome of interest are complex, including interactions between chemicals and non-linear associations (Bobb et al., 2015). In Chapter 3, we found that most phthalates/replacements were associated with urinary hormones in the same direction, and associations were generally linear. Therefore, a future I-KIDS study should first consider using WQSR to evaluate associations of a phthalate/replacement mixture with urinary hormones. However, it will also be important to verify results from WQSR models using other available mixtures approaches. 136 5.2.4. Examine associations of phthalates/replacements with total GWG and trimester-specific GWG In Chapter 4, we evaluated GWG through late pregnancy as our outcome of interest, and we were able to take advantage of available reference charts to calculate GWG z-scores (203, 228, 229). We did not have access to medical records at the time of the study, so we relied on self-reported weights from the last obstetric visit before delivery or those from an earlier obstetric visit within the third trimester. Because of this, we were unable to assess GWG using the IOM categories due to potential for misclassification bias related to not having the final weight in pregnancy. The IOM categories are used by clinicians to provide women with weight gain recommendations based on their pre-pregnancy BMI to support maternal and child health and maximize favorable pregnancy and birth outcomes (3). Often, pregnancy cohort studies assess GWG as an outcome using IOM categories: insufficient, adequate, and excessive. The strength of assessing GWG using IOM categories is that studies evaluating risk factors of GWG can provide clinically-relevant interpretations of their findings. Therefore, evaluating associations between phthalates/replacements and total GWG (using IOM categories) would provide additional context for the results presented in Chapter 4. Future studies in I-KIDS will be able to utilize medical records to ascertain the final pregnancy weight and calculate total GWG. Weight is not gained linearly across pregnancy – it is trimester-specific and follows a sigmoidal curve. Pregnancy is a time of positive energy balance, but energy needs are trimester-specific. First trimester energy needs are the same as before pregnancy leading to very little (if any) weight gained in this trimester (235). Around 95% of weight is gained 137 in the second and third trimesters of pregnancy, with a mean rate of 0.42 kg per week for women classified as normal weight before pregnancy (3, 235). Approaches that summarize GWG into a single measure reduce the ability to evaluate trimester-specific patterns of weight gain and the potential for identifying susceptible windows (211). In I- KIDS, women completed a questionnaire at baseline along with follow-up surveys five more times approximately monthly across gestation and right after delivery. In this questionnaire, women were asked to self-report their measured weight at their most recent obstetric visit. Future studies may consider using these serial weight gain data to calculate the rate of pregnancy weight gain within each trimester and consider associations of phthalates/replacements with GWG trajectories (211). Mixed modeling approaches can also be utilized to evaluate associations of phthalates/replacements with GWG across pregnancy, but also identify if associations are gestational timepoint-specific (211). Once these data are made available, there are many opportunities in I-KIDS to be able to further characterize GWG and understand how phthalates/replacements may impact GWG. 5.2.5. Identify potential metabolic targets of phthalates/replacements in pregnancy As discussed throughout this dissertation, phthalates/replacements are also considered metabolism disrupting chemicals (23), which is important to note given that GWG is regulated by metabolic adaptations in glucose and lipid metabolism (3, 236). Therefore, the next step for understanding associations of phthalates/replacements with GWG may be to investigate whether phthalates/replacements are associated with altered maternal metabolic biomarker concentrations. We have already preliminarily evaluated these 138 associations in I-KIDS - published as an abstract for the International Society of Environmental Epidemiology 2022 Conference (237). Briefly, plasma aliquots of the 13- 22 week fasting blood sample were analyzed for a large panel of maternal metabolic biomarkers, including glucose, insulin, connecting peptide, leptin, free fatty acids, total triglycerides, total cholesterol, and high-density lipoprotein (HDL) cholesterol. Low- density lipoprotein (LDL) and very-low-density lipoprotein (VLDL) levels were calculated using published equations (238). To identify patterns among the metabolic biomarkers, we utilized PCA to reduce ten maternal metabolic factors into three uncorrelated PCs, which explained 73% of the variability in metabolic biomarker concentrations. PC 1 strongly loaded on total, HDL, and LDL cholesterols (cholesterol PC), PC 2 strongly loaded on glucose, insulin, C-peptide, and leptin (glucose homeostasis PC), and PC 3 strongly loaded on triglycerides, VLDL cholesterol, and free fatty acids (lipids PC). We defined strong loading as r > 0.4, and all correlations were positive. Using covariate- adjusted linear regression models, we evaluated associations of individual phthalates/replacements with each of the metabolic PCs. Overall, we observed that DiNP and DEHTP were positively associated with glucose homeostasis PC scores, while DEHP and MEP were positively and negatively associated with lipids PC scores, respectively (238). These findings suggest that concentrations of select phthalate biomarkers measured across pregnancy, including plasticizer replacement DEHTP, are associated with disrupted second-trimester maternal glucose and lipid homeostasis. These critical findings will need to be followed up in the full I-KIDS cohort and will incorporate several additional metabolic, but also inflammatory, biomarkers that are now available. 139 5.2.6. Assess dietary predictors of phthalates/replacements in pregnancy Diet is one of the major exposure sources to phthalates/replacements, particularly plasticizers phthalates since they are used in food processing, and are also found in food packaging materials and the outer coating of medications and supplements (15). Therefore, characterizing dietary sources of phthalates/replacements is critical as recommendations are needed to minimize exposure while concurrently providing pregnant women with accessible and nutritious foods necessary to sustain a healthy pregnancy. In Chapter 2, we evaluated and identified AHEI-2010 – a diet quality index commonly used to characterize whole diets – as an important determinant of phthalate/replacement mixtures. Diet quality indices, such as the AHEI-2010, are excellent proxies for lifestyle behaviors that result in consuming foods with phthalates/replacements since generally unhealthy lifestyles are associated with higher exposure to these chemicals. However, understanding the components of diets that may lead to phthalate exposure, especially exposure to understudied phthalaste replacements DiNCH and DEHTP, is also necessary. This includes conducting additional food- monitoring studies to evaluate the safety of food and dietary patterns (239), but also using existing cohort studies to identify major dietary determinants of gestational urinary phthalate/replacement biomarker concentrations. We published a review of 10 pregnancy cohort studies evaluating dietary predictors of phthalate and bisphenol exposures in pregnancy (240), which is presented in Appendix A. In agreement with prior food- monitoring studies, the use of plastic containers was associated with higher urinary 140 phthalate metabolite concentrations, and increased consumption of canned foods was associated with higher urinary bisphenol A (BPA) concentrations – another prevalent endocrine disrupting chemical. Additionally, foods and dietary patterns associated with a healthier lifestyle, such as organic foods, grown/raised/caught foods, vegetarianism, and folic acid supplementation, as well as some other dietary patterns and foods, including soups and bouillon, spices, and grains, were generally associated with lower urinary phthalate metabolite and bisphenol concentrations in pregnant women. However, not all pregnancy cohort studies were able to reliably detect associations of specific foods/food groups with phthalates and BPA that have been identified as known sources of these chemicals in food-monitoring studies. Therefore, additional well-designed studies are warranted to address these limitations. In I-KIDS, women completed extensive semi- quantitative FFQs at 8-15 and 32-40 weeks gestation that collected information about dietary intakes throughout pregnancy. Future studies in this sample of women will need to verify already known dietary sources of phthalates, but also identify additional dietary determinants of these chemical along with their understudied replacements. 5.2.7. Examine phthalates/replacements as mediators for associations of diet with gestational weight gain and maternal gestational metabolic biomarkers Maternal diet is an established modulator of pregnancy health, as well as fetal growth and development (241) by interacting with inflammatory and metabolic pathways (242-244). Diet is a known modulator of chronic inflammation and oxidative stress, and pregnant women with healthier dietary patterns consume more anti-inflammatory and antioxidant- rich foods that protect against adverse birth and child outcomes (244). Diet also has a 141 major influence on metabolic pathways that are closely linked with inflammation, especially glucose and lipid homeostasis. Pregnant women with healthier dietary patterns tend to have more favorable gestational weight gain, body fat distribution, and metabolic profiles (i.e., levels of insulin, total cholesterol), which are also important determinants of fetal growth and development. Unfortunately, while healthy diets support pregnancy and fetal health, diet is also an established major source of phthalates (discussed previously in section 5.2.5). As presented in section 5.2.4, our preliminary studies observed that select phthalates, particularly replacement DEHTP, were associated with worse mid- pregnancy glucose homeostasis (237). In a recent preliminary study, we also confirmed that higher AHEI-2010 scores, indicating better diet qualities, were associated with favorable mid-pregnancy glucose homeostasis and lower urinary DEHTP metabolite concentrations (data not shown). Therefore, we hypothesize that the favorable metabolic effects of better maternal diet quality may be explained by lower exposure to DEHTP. We aim to conduct a more formal study in the future to test this hypothesis, and also assess GWG as a critical metabolic endpoint in pregnancy. With this study, we hope to show that the maternal metabolic consequences of a poor diet is partly explained by the chemicals that travel with the food supply. 5.2.8. Assess the implications of phthalate/replacement exposure for birth outcomes Our findings in Chapter 4 that a mixture of phthalates/replacements was associated with lower GWG z-scores, which was driven by women carrying females, are interesting given that it is the opposite of what we hypothesized and what many other studies have 142 reported. This is because phthalates are generally considered to be obesogens (23). However, a few studies have corroborated our results, which warrants further investigation to explain these findings (54, 192). It is important to view GWG as a complex phenotype with contributions from the fetus, placenta, amniotic fluid, maternal fluid (i.e., blood, extracellular fluid), protein and fat storage, as well as uterine and breast tissues (198). Our finding that phthalates/replacements were associated with lower GWG, especially in women carrying females, supports the idea that some of these chemicals may target biological pathways that prevent weight gain in one or more of these storage depots (3). Prior experimental and human epidemiologic studies have shown that phthalates are associated with poor birth outcomes, particularly lower birth weight (26, 34-39, 245, 246). Therefore, we hypothesize that associations of phthalates/replacements with lower GWG z-scores in women carrying females may be explained by associations of these chemicals with birth weight. In I-KIDS, we published on associations of parabens, a different class of endocrine disrupting chemicals, with birth size measures (247), which is presented in Appendix B. We observed that maternal urinary methylparaben and propylparaben concentrations were negatively associated with birth weight, birth weight z-scores, body length, and weight/length ratio in female, but not male newborns; these results persisted even after additionally adjusting for gestational length. Given that parabens and phthalates have been shown to have similar biological targets, phthalates/replacements may also be associated with lower female birth size. Therefore, a future study is needed to investigate this to further explain our observed associations between phthalates/replacements and GWG z-scores, but also 143 begin to understand the impact of these chemicals on child health outcomes within the context of maternal health disruption. 5.2.9. Determine if associations of phthalates/replacement with pregnancy endocrine and metabolic-related endpoints persist beyond pregnancy As discussed throughout this dissertation, prenatal phthalate/replacement exposure is associated with adverse pregnancy/birth outcomes, long-term child health outcomes, but also may have long-term repercussions on maternal health (2). Several recent reviews suggest that various pregnancy pathologies, as well as the act of being pregnant, may be a “stress test” – in that pregnancy may serve as a first glance into potential long-term health outcomes in women (2). Additionally, in one of our published studies using data from the Midlife Women’s Health Study (248), we showed that pregnancy history, including age at first birth and parity, were important determinants of maternal midlife metabolic health (presented in Appendix C), further supporting pregnancy as a critical window for women’s later life health. In the same midlife cohort, we evaluated cross- sectional associations of midlife phthalate exposure with endocrine-related endpoints (presented in Appendixes D, E, F). We observed that higher concentrations of select urinary phthalate metabolites were associated with higher estradiol, testosterone, progesterone, and anti-Mullerian hormone concentrations (249), with higher odds of experiencing hot flashes in the past 30 days and experiencing daily/weekly hot flashes (250), and with higher risk of having a prior fibroid diagnosis (251). However, the relevant etiologic window of exposure to endocrine disrupting chemicals that may cause these midlife endpoints likely happens well before midlife, and potentially during pregnancy. 144 One likely mechanism behind this hypothesis is that exposure to phthalates/replacements in pregnancy alters maternal hormones (as we showed in Chapter 3), which impacts women’s health long after pregnancy. For example, a prospective case-control study found that higher gestational estrogen concentrations were associated with increased risk of breast cancer in mothers after 38 years of follow-up (168). Changes in estrogens, as well as androgens, may also be implicated in cardiovascular disease and osteoporosis, which are prevalent in post-menopausal women (252). Experimental studies and those in pregnancy cohorts have begun trying to establish the link between gestational chemical exposures and women’s health after pregnancy and have shown that maternal phthalate exposure during pregnancy is associated with postpartum weight retention that may persist six years after delivery (53, 54, 253). This will also be evaluated using data from I-KIDS women in a new study called the Illinois Metabolic Outcomes in Moms (I-MOMS) where at least 350 I-KIDS women will be recruited four-to-seven years after their I-KIDS pregnancy. This future study will provide critical findings about the long-term implications of association between phthalates/replacement and maternal lifelong cardiometabolic and endocrine health. The findings presented in this dissertation provided important evidence about the potential maternal health consequences of phthalate/replacement exposure in pregnancy, which is an important foundation for addressing these long-term goals. 145 BIBLIOGRAPHY 1. Livingston G. They're Waiting Longer, but U.S. Women Today More Likely to Have Children than a Decade Ago. Pew Research Center, 2018. 2. Haggerty DK, Upson K, Pacyga DC, Franko JE, Braun JM, Strakovsky RS. REPRODUCTIVE TOXICOLOGY: Pregnancy exposure to endocrine disrupting chemicals: implications for women's health. Reproduction. 2021;162(5):F169-F80. doi: 10.1530/REP-21-0051. PubMed PMID: 34486984; PMCID: PMC8511181. 3. Rasmussen KM, Yaktine AL. Weight Gain During Pregnancy: Reexamining the Guidelines. Washington (DC)2009. 4. Tal R, Taylor HS, Burney RO, Mooney SB, Giudice LC. Endocrinology of Pregnancy. In: Feingold KR, Anawalt B, Boyce A, Chrousos G, Dungan K, Grossman A, Hershman JM, Kaltsas G, Koch C, Kopp P, Korbonits M, McLachlan R, Morley JE, New M, Perreault L, Purnell J, Rebar R, Singer F, Trence DL, Vinik A, Wilson DP, editors. Endotext. South Dartmouth (MA)2000. 5. Mor G, Aldo P, Alvero AB. The unique immunological and microbial aspects of pregnancy. Nat Rev Immunol. 2017;17(8):469-82. doi: 10.1038/nri.2017.64. PubMed PMID: 28627518. 6. Schock H, Zeleniuch-Jacquotte A, Lundin E, Grankvist K, Lakso HA, Idahl A, Lehtinen M, Surcel HM, Fortner RT. Hormone concentrations throughout uncomplicated pregnancies: a longitudinal study. BMC Pregnancy Childbirth. 2016;16(1):146. doi: 10.1186/s12884-016-0937-5. PubMed PMID: 27377060; PMCID: PMC4932669. 7. Makieva S, Saunders PT, Norman JE. Androgens in pregnancy: roles in parturition. Hum Reprod Update. 2014;20(4):542-59. Epub 2014/03/20. doi: 10.1093/humupd/dmu008. PubMed PMID: 24643344; PMCID: PMC4063701. 8. Morisset AS, Dube MC, Drolet R, Pelletier M, Labrie F, Luu-The V, Tremblay Y, Robitaille J, John Weisnagel S, Tchernof A. Androgens in the maternal and fetal circulation: association with insulin resistance. J Matern Fetal Neonatal Med. 2013;26(5):513-9. doi: 10.3109/14767058.2012.735725. PubMed PMID: 23075231. 9. Salamalekis E, Bakas P, Vitoratos N, Eleptheriadis M, Creatsas G. Androgen levels in the third trimester of pregnancy in patients with preeclampsia. Eur J Obstet Gynecol Reprod Biol. 2006;126(1):16-9. doi: 10.1016/j.ejogrb.2005.07.007. PubMed PMID: 16139944. 10. Gluckman PDB, T.; Hanson, M.A. The developmental origins of health and disease (DOHaD) concept: past, present, and future. Press A, editor. Washington, DC, USA2016. 1-15 p. 146 11. Dassanayake M, Langen E, Davis MB. Pregnancy Complications as a Window to Future Cardiovascular Disease. Cardiology in review. 2020;28(1):14-9. Epub 2019/04/23. doi: 10.1097/crd.0000000000000253. PubMed PMID: 31008769. 12. Hales CN, Barker DJ, Clark PM, Cox LJ, Fall C, Osmond C, Winter PD. Fetal and infant growth and impaired glucose tolerance at age 64. BMJ. 1991;303(6809):1019-22. Epub 1991/10/26. doi: 10.1136/bmj.303.6809.1019. PubMed PMID: 1954451; PMCID: PMC1671766. 13. Boyles AL, Beverly BE, Fenton SE, Jackson CL, Jukic AMZ, Sutherland VL, Baird DD, Collman GW, Dixon D, Ferguson KK, Hall JE, Martin EM, Schug TT, White AJ, Chandler KJ. Environmental Factors Involved in Maternal Morbidity and Mortality. Journal of women's health (2002). 2020. Epub 2020/11/20. doi: 10.1089/jwh.2020.8855. PubMed PMID: 33211615. 14. Woodruff TJ, Zota AR, Schwartz JM. Environmental chemicals in pregnant women in the United States: NHANES 2003-2004. Environ Health Perspect. 2011;119(6):878-85. Epub 2011/01/15. doi: 10.1289/ehp.1002727. PubMed PMID: 21233055; PMCID: PMC3114826. 15. CDC. Biomonitoring Summary: Di-2-ethylhexyl Phthalate 2017 [cited 2023]. Available from: https://www.cdc.gov/biomonitoring/DEHP_BiomonitoringSummary.html. 16. CDC. Biomonitoring Summary: Diethyl Phthalate 2017 [cited 2023]. Available from: https://www.cdc.gov/biomonitoring/DEP_BiomonitoringSummary.html. 17. James-Todd TM, Meeker JD, Huang T, Hauser R, Seely EW, Ferguson KK, Rich-Edwards JW, McElrath TF. Racial and ethnic variations in phthalate metabolite concentration changes across full-term pregnancies. J Expo Sci Environ Epidemiol. 2017;27(2):160-6. doi: 10.1038/jes.2016.2. PubMed PMID: 26860587; PMCID: PMC4980273. 18. Wenzel AG, Brock JW, Cruze L, Newman RB, Unal ER, Wolf BJ, Somerville SE, Kucklick JR. Prevalence and predictors of phthalate exposure in pregnant women in Charleston, SC. Chemosphere. 2018;193:394-402. doi: 10.1016/j.chemosphere.2017.11.019. PubMed PMID: 29154114; PMCID: PMC6282186. 19. Harris CA, Henttu P, Parker MG, Sumpter JP. The estrogenic activity of phthalate esters in vitro. Environ Health Perspect. 1997;105(8):802-11. doi: 10.1289/ehp.97105802. PubMed PMID: 9347895; PMCID: PMC1470189. 20. Jobling S, Reynolds T, White R, Parker MG, Sumpter JP. A variety of environmentally persistent chemicals, including some phthalate plasticizers, are weakly estrogenic. Environ Health Perspect. 1995;103(6):582-7. doi: 10.1289/ehp.95103582. PubMed PMID: 7556011; PMCID: PMC1519124. 147 21. Howdeshell KL, Furr J, Lambright CR, Rider CV, Wilson VS, Gray LE, Jr. Cumulative effects of dibutyl phthalate and diethylhexyl phthalate on male rat reproductive tract development: altered fetal steroid hormones and genes. Toxicol Sci. 2007;99(1):190-202. doi: 10.1093/toxsci/kfm069. PubMed PMID: 17400582. 22. Parks LG, Ostby JS, Lambright CR, Abbott BD, Klinefelter GR, Barlow NJ, Gray LE, Jr. The plasticizer diethylhexyl phthalate induces malformations by decreasing fetal testosterone synthesis during sexual differentiation in the male rat. Toxicol Sci. 2000;58(2):339-49. doi: 10.1093/toxsci/58.2.339. PubMed PMID: 11099646. 23. Veiga-Lopez A, Pu Y, Gingrich J, Padmanabhan V. Obesogenic Endocrine Disrupting Chemicals: Identifying Knowledge Gaps. Trends Endocrinol Metab. 2018;29(9):607-25. Epub 2018/07/19. doi: 10.1016/j.tem.2018.06.003. PubMed PMID: 30017741; PMCID: PMC6098722. 24. Toft G, Jonsson BA, Lindh CH, Jensen TK, Hjollund NH, Vested A, Bonde JP. Association between pregnancy loss and urinary phthalate levels around the time of conception. Environ Health Perspect. 2012;120(3):458-63. Epub 2011/11/25. doi: 10.1289/ehp.1103552. PubMed PMID: 22113848; PMCID: PMC3295336. 25. Cantonwine DE, Meeker JD, Ferguson KK, Mukherjee B, Hauser R, McElrath TF. Urinary Concentrations of Bisphenol A and Phthalate Metabolites Measured during Pregnancy and Risk of Preeclampsia. Environmental health perspectives. 2016;124(10):1651-5. Epub 2016/05/13. doi: 10.1289/EHP188. PubMed PMID: 27177253. 26. Ferguson KK, McElrath TF, Cantonwine DE, Mukherjee B, Meeker JD. Phthalate metabolites and bisphenol-A in association with circulating angiogenic biomarkers across pregnancy. Placenta. 2015;36(6):699-703. Epub 2015/04/14. doi: 10.1016/j.placenta.2015.04.002. PubMed PMID: 25913709. 27. Yan D, Jiao Y, Yan H, Liu T, Yan H, Yuan J. Endocrine-disrupting chemicals and the risk of gestational diabetes mellitus: a systematic review and meta-analysis. Environ Health. 2022;21(1):53. Epub 2022/05/17. doi: 10.1186/s12940-022-00858-8. PubMed PMID: 35578291; PMCID: PMC9109392. 28. Adibi JJ, Hauser R, Williams PL, Whyatt RM, Calafat AM, Nelson H, Herrick R, Swan SH. Maternal urinary metabolites of Di-(2-Ethylhexyl) phthalate in relation to the timing of labor in a US multicenter pregnancy cohort study. Am J Epidemiol. 2009;169(8):1015-24. Epub 2009/03/03. doi: 10.1093/aje/kwp001. PubMed PMID: 19251754; PMCID: PMC2727228. 29. Ferguson KK, McElrath TF, Meeker JD. Environmental phthalate exposure and preterm birth. JAMA Pediatr. 2014;168(1):61-7. Epub 2013/11/20. doi: 10.1001/jamapediatrics.2013.3699. PubMed PMID: 24247736; PMCID: PMC4005250. 30. Welch BM, Keil AP, Buckley JP, Calafat AM, Christenbury KE, Engel SM, O'Brien KM, Rosen EM, James-Todd T, Zota AR, Ferguson KK, Pooled Phthalate E, Preterm 148 Birth Study G, Alshawabkeh AN, Cordero JF, Meeker JD, Barrett ES, Bush NR, Nguyen RHN, Sathyanarayana S, Swan SH, Cantonwine DE, McElrath TF, Aalborg J, Dabelea D, Starling AP, Hauser R, Messerlian C, Zhang Y, Bradman A, Eskenazi B, Harley KG, Holland N, Bloom MS, Newman RB, Wenzel AG, Braun JM, Lanphear BP, Yolton K, Factor-Litvak P, Herbstman JB, Rauh VA, Drobnis EZ, Sparks AE, Redmon JB, Wang C, Binder AM, Michels KB, Baird DD, Jukic AMZ, Weinberg CR, Wilcox AJ, Rich DQ, Weinberger B, Padmanabhan V, Watkins DJ, Hertz-Picciotto I, Schmidt RJ. Associations Between Prenatal Urinary Biomarkers of Phthalate Exposure and Preterm Birth: A Pooled Study of 16 US Cohorts. JAMA Pediatr. 2022;176(9):895-905. Epub 2022/07/12. doi: 10.1001/jamapediatrics.2022.2252. PubMed PMID: 35816333; PMCID: PMC9274448 Institutes of Health (NIH)/National Institute of Environmental Health Sciences (NIEHS) and the US Environmental Protection Agency (EPA) during the conduct of the study; honorarium for grant review from the NIH/Center for Scientific Review outside the submitted work; and honorarium for advisory board participation from the University of Montana outside the submitted work. Dr Cordero reported grants from the NIH during the conduct of the study and outside the submitted work and from Medtronic Foundation outside the submitted work. Dr Barrett reported grants from NIH during the conduct of the study. Dr Bush reported grants from the NIH during the conduct of the study. Dr McElrath reported research support to their institution and equity from NxPrenatal Inc; serving on the scientific advisory board of and equity from Mirvie Inc; and serving on the scientific advisory board of and cash payment from Hoffmann-La Roche, Momenta Pharmaceuticals, Comanche Biopharma; and Tectonic Therapeutic. Dr Starling reported grants from the NIH during the conduct of the study. Dr Hauser reported grants from NIEHS during the conduct of the study. Dr Eskenazi reported grants from the NIH and EPA during the conduct of the study. Dr Harley reported grants from the NIEHS during the conduct of the study. Dr Holland reported grants from the NIEHS during the conduct of the study. Dr Bloom reported grants from the NIH during the conduct of the study. Dr Braun reported grants from the NIH during the conduct of the study and served as an expert witness for plaintiffs in litigation related to perfluoroalkyl substances-contaminated drinking water for Morgan & Morgan Law Firm (funds were not paid to Dr Braun directly; all compensation was paid to a discretionary account that cannot be used for salary or fringe) outside the submitted work. Dr Factor-Litvak reported grants from the NIH during the conduct of the study. Dr Jukic reported grants from the NIEHS during the conduct of the study. Dr Weinberg reported salary support from the NIEHS during the conduct of the study. Dr Weinberger reported grants from the NIH and the New Jersey Department of Environmental Protection during the conduct of the study. Dr Watkins reported grants from the NIH and EPA during the conduct of the study. Dr Schmidt reported grants from Autism Science Foundation during the conduct of the study. No other disclosures were reported. 31. Gao H, Zhu BB, Huang K, Zhu YD, Yan SQ, Wu XY, Han Y, Sheng J, Cao H, Zhu P, Tao FB. Effects of single and combined gestational phthalate exposure on blood pressure, blood glucose and gestational weight gain: A longitudinal analysis. Environ Int. 2021;155:106677. Epub 2021/06/15. doi: 10.1016/j.envint.2021.106677. PubMed PMID: 34126297. 149 32. James-Todd TM, Meeker JD, Huang T, Hauser R, Ferguson KK, Rich-Edwards JW, McElrath TF, Seely EW. Pregnancy urinary phthalate metabolite concentrations and gestational diabetes risk factors. Environment International. 2016;96:118-26. doi: https://doi.org/10.1016/j.envint.2016.09.009. 33. Li J, Qian X, Zhao H, Zhou Y, Xu S, Li Y, Xiang L, Shi J, Xia W, Cai Z. Determinants of exposure levels, metabolism, and health risks of phthalates among pregnant women in Wuhan, China. Ecotoxicol Environ Saf. 2019;184:109657. doi: 10.1016/j.ecoenv.2019.109657. PubMed PMID: 31526923. 34. Tanaka T. Reproductive and neurobehavioural effects of bis(2-ethylhexyl) phthalate (DEHP) in a cross-mating toxicity study of mice. Food Chem Toxicol. 2005;43(4):581-9. Epub 2005/02/22. doi: 10.1016/j.fct.2005.01.001. PubMed PMID: 15721206. 35. Boberg J, Metzdorff S, Wortziger R, Axelstad M, Brokken L, Vinggaard AM, Dalgaard M, Nellemann C. Impact of diisobutyl phthalate and other PPAR agonists on steroidogenesis and plasma insulin and leptin levels in fetal rats. Toxicology. 2008;250(2-3):75-81. Epub 2008/07/08. doi: 10.1016/j.tox.2008.05.020. PubMed PMID: 18602967. 36. Tyl RW, Myers CB, Marr MC, Fail PA, Seely JC, Brine DR, Barter RA, Butala JH. Reproductive toxicity evaluation of dietary butyl benzyl phthalate (BBP) in rats. Reprod Toxicol. 2004;18(2):241-64. Epub 2004/03/17. doi: 10.1016/j.reprotox.2003.10.006. PubMed PMID: 15019722. 37. Gray LE, Jr., Wilson VS, Stoker T, Lambright C, Furr J, Noriega N, Howdeshell K, Ankley GT, Guillette L. Adverse effects of environmental antiandrogens and androgens on reproductive development in mammals. Int J Androl. 2006;29(1):96-104; discussion 5-8. Epub 2006/02/10. doi: 10.1111/j.1365-2605.2005.00636.x. PubMed PMID: 16466529. 38. Weaver JA, Beverly BEJ, Keshava N, Mudipalli A, Arzuaga X, Cai C, Hotchkiss AK, Makris SL, Yost EE. Hazards of diethyl phthalate (DEP) exposure: A systematic review of animal toxicology studies. Environ Int. 2020;145:105848. doi: 10.1016/j.envint.2020.105848. PubMed PMID: 32958228; PMCID: PMC7995140. 39. Neier K, Cheatham D, Bedrosian LD, Dolinoy DC. Perinatal exposures to phthalates and phthalate mixtures result in sex-specific effects on body weight, organ weights and intracisternal A-particle (IAP) DNA methylation in weanling mice. J Dev Orig Health Dis. 2019;10(2):176-87. doi: 10.1017/S2040174418000430. PubMed PMID: 29991372; PMCID: PMC6329673. 40. Li N, Liu E, Guo J, Pan L, Li B, Wang P, Liu J, Wang Y, Liu G, Baccarelli AA, Hou L, Hu G. Maternal prepregnancy body mass index and gestational weight gain on pregnancy outcomes. PLoS One. 2013;8(12):e82310. doi: 10.1371/journal.pone.0082310. PubMed PMID: 24376527. 150 41. Caulfield LE, Witter FR, Stoltzfus RJ. Determinants of gestational weight gain outside the recommended ranges among black and white women. Obstet Gynecol. 1996;87(5 Pt 1):760-6. doi: 10.1016/0029-7844(96)00023-3. PubMed PMID: 8677082. 42. Wu Y, Wan S, Gu S, Mou Z, Dong L, Luo Z, Zhang J, Hua X. Gestational weight gain and adverse pregnancy outcomes: a prospective cohort study. BMJ Open. 2020;10(9):e038187. doi: 10.1136/bmjopen-2020-038187. PubMed PMID: 32878761; PMCID: PMC7470642. 43. Zanardo V, Giliberti L, Giliberti E, Grassi A, Perin V, Parotto M, Soldera G, Straface G. The role of gestational weight gain disorders in symptoms of maternal postpartum depression. Int J Gynaecol Obstet. 2021;153(2):234-8. doi: 10.1002/ijgo.13445. PubMed PMID: 33113162. 44. Lau EY, Liu J, Archer E, McDonald SM, Liu J. Maternal weight gain in pregnancy and risk of obesity among offspring: a systematic review. J Obes. 2014;2014:524939. doi: 10.1155/2014/524939. PubMed PMID: 25371815; PMCID: PMC4202338. 45. Ipapo KN, Factor-Litvak P, Whyatt RM, Calafat AM, Diaz D, Perera F, Rauh V, Herbstman JB. Maternal prenatal urinary phthalate metabolite concentrations and visual recognition memory among infants at 27 weeks. Environ Res. 2017;155:7-14. Epub 2017/02/09. doi: 10.1016/j.envres.2017.01.019. PubMed PMID: 28171772; PMCID: PMC5366271. 46. Philippat C, Nakiwala D, Calafat AM, Botton J, De Agostini M, Heude B, Slama R, Group EM-CS. Prenatal Exposure to Nonpersistent Endocrine Disruptors and Behavior in Boys at 3 and 5 Years. Environ Health Perspect. 2017;125(9):097014. Epub 2017/09/25. doi: 10.1289/EHP1314. PubMed PMID: 28937960; PMCID: PMC5915182. 47. Jurewicz J, Hanke W. Exposure to phthalates: reproductive outcome and children health. A review of epidemiological studies. Int J Occup Med Environ Health. 2011;24(2):115-41. Epub 2011/05/20. doi: 10.2478/s13382-011-0022-2. PubMed PMID: 21594692. 48. Gascon M, Casas M, Morales E, Valvi D, Ballesteros-Gomez A, Luque N, Rubio S, Monfort N, Ventura R, Martinez D, Sunyer J, Vrijheid M. Prenatal exposure to bisphenol A and phthalates and childhood respiratory tract infections and allergy. J Allergy Clin Immunol. 2015;135(2):370-8. Epub 2014/12/03. doi: 10.1016/j.jaci.2014.09.030. PubMed PMID: 25445825. 49. Vernet C, Pin I, Giorgis-Allemand L, Philippat C, Benmerad M, Quentin J, Calafat AM, Ye X, Annesi-Maesano I, Siroux V, Slama R, Group EM-CCS. In Utero Exposure to Select Phenols and Phthalates and Respiratory Health in Five-Year-Old Boys: A Prospective Study. Environ Health Perspect. 2017;125(9):097006. Epub 2017/09/22. doi: 10.1289/EHP1015. PubMed PMID: 28934727; PMCID: PMC5915196. 50. Harley KG, Berger K, Rauch S, Kogut K, Claus Henn B, Calafat AM, Huen K, Eskenazi B, Holland N. Association of prenatal urinary phthalate metabolite 151 concentrations and childhood BMI and obesity. Pediatr Res. 2017;82(3):405-15. Epub 2017/04/21. doi: 10.1038/pr.2017.112. PubMed PMID: 28426647; PMCID: PMC5581502. 51. Martinez-Martinez MI, Alegre-Martinez A, Cauli O. Prenatal exposure to phthalates and its effects upon cognitive and motor functions: A systematic review. Toxicology. 2021;463:152980. Epub 2021/10/09. doi: 10.1016/j.tox.2021.152980. PubMed PMID: 34624397. 52. Golestanzadeh M, Riahi R, Kelishadi R. Association of exposure to phthalates with cardiometabolic risk factors in children and adolescents: a systematic review and meta-analysis. Environ Sci Pollut Res Int. 2019;26(35):35670-86. Epub 2019/11/16. doi: 10.1007/s11356-019-06589-7. PubMed PMID: 31728953. 53. Philips EM, Jaddoe VWV, Deierlein A, Asimakopoulos AG, Kannan K, Steegers EAP, Trasande L. Exposures to phthalates and bisphenols in pregnancy and postpartum weight gain in a population-based longitudinal birth cohort. Environment International. 2020;144:106002. doi: https://doi.org/10.1016/j.envint.2020.106002. 54. Deierlein AL, Wu H, Just AC, Kupsco AJ, Braun JM, Oken E, Soria-Contreras DC, Cantoral A, Pizano ML, McRae N, Tellez-Rojo MM, Wright RO, Baccarelli AA. Prenatal phthalates, gestational weight gain, and long-term weight changes among Mexican women. Environ Res. 2022;209:112835. Epub 2022/02/02. doi: 10.1016/j.envres.2022.112835. PubMed PMID: 35101400; PMCID: PMC8976769. 55. Wu H, Kupsco A, Just A, Calafat AM, Oken E, Braun JM, Sanders AP, Mercado- Garcia A, Cantoral A, Pantic I, Tellez-Rojo MM, Wright RO, Baccarelli AA, Deierlein AL. Maternal Phthalates Exposure and Blood Pressure during and after Pregnancy in the PROGRESS Study. Environ Health Perspect. 2021;129(12):127007. Epub 2021/12/23. doi: 10.1289/EHP8562. PubMed PMID: 34935432; PMCID: PMC8693773. 56. Jacobson MH, Stein CR, Liu M, Ackerman MG, Blakemore JK, Long SE, Pinna G, Romay-Tallon R, Kannan K, Zhu H, Trasande L. Prenatal Exposure to Bisphenols and Phthalates and Postpartum Depression: The Role of Neurosteroid Hormone Disruption. J Clin Endocrinol Metab. 2021;106(7):1887-99. Epub 2021/04/02. doi: 10.1210/clinem/dgab199. PubMed PMID: 33792735; PMCID: PMC8502446. 57. Hu L, Mei H, Feng H, Huang Y, Cai X, Xiang F, Chen L, Xiao H. Exposure to bisphenols, parabens and phthalates during pregnancy and postpartum anxiety and depression symptoms: Evidence from women with twin pregnancies. Environ Res. 2023;221:115248. Epub 2023/01/10. doi: 10.1016/j.envres.2023.115248. PubMed PMID: 36623682. 58. Kamrin MA. Phthalate risks, phthalate regulation, and public health: a review. Journal of toxicology and environmental health Part B, Critical reviews. 2009;12(2):157- 74. Epub 2009/02/25. doi: 10.1080/10937400902729226. PubMed PMID: 19235623. 152 59. Wadey BL. An innovative plasticizer for sensitive applications. Journal of Vinyl & Additive Technology. 2003;9(4):172-6. doi: 10.1002/vnl.10080. PubMed PMID: WOS:000187952100004. 60. Abe Y, Yamaguchi M, Mutsuga M, Hirahara Y, Kawamura Y. Survey of Plasticizers in Polyvinyl Chloride Toys. Food Hyg Saf Sci. 2012;53(1):19-27. doi: 10.3358/shokueishi.53.19. PubMed PMID: WOS:000301510200004. 61. McCombie G, Biedermann S, Suter G, Biedermann M. Survey on plasticizers currently found in PVC toys on the Swiss market: Banned phthalates are only a minor concern. J Environ Sci Health A Tox Hazard Subst Environ Eng. 2017;52(5):491-6. Epub 2017/01/28. doi: 10.1080/10934529.2016.1274176. PubMed PMID: 28129041. 62. Yland JJ, Zhang Y, Williams PL, Mustieles V, Vagios S, Souter I, Calafat AM, Hauser R, Messerlian C, Earth Study T. Phthalate and DINCH urinary concentrations across pregnancy and risk of preterm birth. Environ Pollut. 2022;292(Pt B):118476. doi: 10.1016/j.envpol.2021.118476. PubMed PMID: 34763012. 63. Maertens A, Golden E, Hartung T. Avoiding Regrettable Substitutions: Green Toxicology for Sustainable Chemistry. ACS Sustain Chem Eng. 2021;9(23):7749-58. Epub 2021/06/14. doi: 10.1021/acssuschemeng.0c09435. PubMed PMID: 36051558; PMCID: PMC9432817. 64. CDC. Fourth National Report on Human Exposure to Environmental Chemicals (Updated Tables, March 2021). Atlanta, Georgia: U.S. Department of Health and Human Services, Centers for Disease Control and Prevention, 2021. 65. Lessmann F, Schutze A, Weiss T, Bruning T, Koch HM. Determination of metabolites of di(2-ethylhexyl) terephthalate (DEHTP) in human urine by HPLC-MS/MS with on-line clean-up. J Chromatogr B Analyt Technol Biomed Life Sci. 2016;1011:196- 203. Epub 2016/01/17. doi: 10.1016/j.jchromb.2015.12.042. PubMed PMID: 26773884. 66. Braun JM, Gennings C, Hauser R, Webster TF. What Can Epidemiological Studies Tell Us about the Impact of Chemical Mixtures on Human Health? Environ Health Perspect. 2016;124(1):A6-9. doi: 10.1289/ehp.1510569. PubMed PMID: 26720830. 67. Kortenkamp A, Faust M. Regulate to reduce chemical mixture risk. Science. 2018;361(6399):224-6. doi: 10.1126/science.aat9219. PubMed PMID: 30026211. 68. Hotelling H. Analysis of a complex of statistical variables into principal components. Journal of Educational Psychology. 1933;24(6):417-41. doi: https://doi.org/10.1037/h0071325. 69. MacQueen J. Some methods for classification and analysis of multivariate observations. Proceedings of 5th Berkeley Symposium on Mathematical Statistics and Probability1967. p. 281-97. 153 70. Keil A. qgcomp: Quantile G-Computation. 2022. p. R package version 2.8.6. https://CRAN.R-project.org/package=qgcomp. 71. Ranzetti SC, P.; Just, A.C.; Bello, G.; Gennings, C. gWQS: Generalized Weighted Quantile Sum Regression. 2021. p. R package version 3.0.4. https://CRAN.R- project.org/package=gWQS. 72. Keil AP, Buckley JP, O'Brien KM, Ferguson KK, Zhao S, White AJ. A Quantile- Based g-Computation Approach to Addressing the Effects of Exposure Mixtures. Environ Health Perspect. 2020;128(4):47004. Epub 2020/04/08. doi: 10.1289/EHP5838. PubMed PMID: 32255670; PMCID: PMC7228100. 73. Carrico C, Gennings C, Wheeler DC, Factor-Litvak P. Characterization of Weighted Quantile Sum Regression for Highly Correlated Data in a Risk Analysis Setting. J Agric Biol Environ Stat. 2015;20(1):100-20. doi: 10.1007/s13253-014-0180-3. PubMed PMID: 30505142; PMCID: PMC6261506. 74. Gallicchio L, Schilling C, Romani WA, Miller S, Zacur H, Flaws JA. Endogenous hormones, participant characteristics, and symptoms among midlife women. Maturitas. 2008;59(2):114-27. Epub 2008/03/04. doi: 10.1016/j.maturitas.2008.01.003. PubMed PMID: 18313243; PMCID: PMC2302829. 75. Diamanti-Kandarakis E, Bourguignon JP, Giudice LC, Hauser R, Prins GS, Soto AM, Zoeller RT, Gore AC. Endocrine-disrupting chemicals: an Endocrine Society scientific statement. Endocr Rev. 2009;30(4):293-342. Epub 2009/06/09. doi: 10.1210/er.2009-0002. PubMed PMID: 19502515; PMCID: PMC2726844. 76. Johns LE, Ferguson KK, McElrath TF, Mukherjee B, Meeker JD. Associations between Repeated Measures of Maternal Urinary Phthalate Metabolites and Thyroid Hormone Parameters during Pregnancy. Environ Health Perspect. 2016;124(11):1808- 15. doi: 10.1289/EHP170. PubMed PMID: 27152641; PMCID: PMC5089879 interests. 77. Johns LE, Ferguson KK, Soldin OP, Cantonwine DE, Rivera-Gonzalez LO, Del Toro LV, Calafat AM, Ye X, Alshawabkeh AN, Cordero JF, Meeker JD. Urinary phthalate metabolites in relation to maternal serum thyroid and sex hormone levels during pregnancy: a longitudinal analysis. Reprod Biol Endocrinol. 2015;13:4. Epub 2015/01/19. doi: 10.1186/1477-7827-13-4. PubMed PMID: 25596636; PMCID: PMC4326411. 78. Engel A, Buhrke T, Imber F, Jessel S, Seidel A, Völkel W, Lampen A. Agonistic and antagonistic effects of phthalates and their urinary metabolites on the steroid hormone receptors ERα, ERβ, and AR. Toxicology Letters. 2017;277:54-63. doi: https://doi.org/10.1016/j.toxlet.2017.05.028. 79. Gore AC, Chappell VA, Fenton SE, Flaws JA, Nadal A, Prins GS, Toppari J, Zoeller RT. EDC-2: The Endocrine Society's Second Scientific Statement on Endocrine- Disrupting Chemicals. Endocr Rev. 2015;36(6):E1-E150. doi: 10.1210/er.2015-1010. PubMed PMID: 26544531; PMCID: PMC4702494. 154 80. James-Todd TM, Chiu YH, Messerlian C, Mínguez-Alarcón L, Ford JB, Keller M, Petrozza J, Williams PL, Ye X, Calafat AM, Hauser R. Trimester-specific phthalate concentrations and glucose levels among women from a fertility clinic. Environ Health. 2018;17(1):55. Epub 2018/06/15. doi: 10.1186/s12940-018-0399-5. PubMed PMID: 29898728; PMCID: PMC6000948. 81. Shaffer RM, Ferguson KK, Sheppard L, James-Todd T, Butts S, Chandrasekaran S, Swan SH, Barrett ES, Nguyen R, Bush N, McElrath TF, Sathyanarayana S. Maternal urinary phthalate metabolites in relation to gestational diabetes and glucose intolerance during pregnancy. Environ Int. 2019;123:588-96. Epub 2019/01/10. doi: 10.1016/j.envint.2018.12.021. PubMed PMID: 30622083; PMCID: PMC6347428. 82. Ferguson KK, McElrath TF, Ko YA, Mukherjee B, Meeker JD. Variability in urinary phthalate metabolite levels across pregnancy and sensitive windows of exposure for the risk of preterm birth. Environ Int. 2014;70:118-24. Epub 2014/06/18. doi: 10.1016/j.envint.2014.05.016. PubMed PMID: 24934852; PMCID: PMC4104181. 83. Lee G, Kim S, Bastiaensen M, Malarvannan G, Poma G, Caballero Casero N, Gys C, Covaci A, Lee S, Lim JE, Mok S, Moon HB, Choi G, Choi K. Exposure to organophosphate esters, phthalates, and alternative plasticizers in association with uterine fibroids. Environ Res. 2020;189:109874. doi: 10.1016/j.envres.2020.109874. PubMed PMID: 32678732. 84. Long SE, Kahn LG, Trasande L, Jacobson MH. Urinary phthalate metabolites and alternatives and serum sex steroid hormones among pre- and postmenopausal women from NHANES, 2013-16. Sci Total Environ. 2021;769:144560. doi: 10.1016/j.scitotenv.2020.144560. PubMed PMID: 33493905; PMCID: PMC7969426. 85. Derakhshan A, Shu H, Broeren MAC, Lindh CH, Peeters RP, Kortenkamp A, Demeneix B, Bornehag CG, Korevaar TIM. Association of phthalate exposure with thyroid function during pregnancy. Environ Int. 2021;157:106795. doi: 10.1016/j.envint.2021.106795. PubMed PMID: 34358912. 86. Pacyga DC, Gardiner JC, Flaws JA, Li Z, Calafat AM, Korrick SA, Schantz SL, Strakovsky RS. Maternal phthalate and phthalate alternative metabolites and urinary biomarkers of estrogens and testosterones across pregnancy. Environ Int. 2021;155:106676. doi: 10.1016/j.envint.2021.106676. PubMed PMID: 34116379; PMCID: PMC8292204. 87. van TETJ, Rosen EM, Barrett ES, Nguyen RHN, Sathyanarayana S, Milne GL, Calafat AM, Swan SH, Ferguson KK. Phthalates and Phthalate Alternatives Have Diverse Associations with Oxidative Stress and Inflammation in Pregnant Women. Environ Sci Technol. 2019;53(6):3258-67. doi: 10.1021/acs.est.8b05729. PubMed PMID: 30793895; PMCID: PMC6487641. 88. Shu H, Jonsson BA, Gennings C, Svensson A, Nanberg E, Lindh CH, Knutz M, Takaro TK, Bornehag CG. Temporal trends of phthalate exposures during 2007-2010 in 155 Swedish pregnant women. J Expo Sci Environ Epidemiol. 2018;28(5):437-47. doi: 10.1038/s41370-018-0020-6. PubMed PMID: 29472621. 89. Casas L, Fernandez MF, Llop S, Guxens M, Ballester F, Olea N, Irurzun MB, Rodriguez LS, Riano I, Tardon A, Vrijheid M, Calafat AM, Sunyer J, Project I. Urinary concentrations of phthalates and phenols in a population of Spanish pregnant women and children. Environ Int. 2011;37(5):858-66. doi: 10.1016/j.envint.2011.02.012. PubMed PMID: 21440302. 90. Valvi D, Monfort N, Ventura R, Casas M, Casas L, Sunyer J, Vrijheid M. Variability and predictors of urinary phthalate metabolites in Spanish pregnant women. Int J Hyg Environ Health. 2015;218(2):220-31. Epub 2015/01/07. doi: 10.1016/j.ijheh.2014.11.003. PubMed PMID: 25558797. 91. He X, Zang J, Liao P, Zheng Y, Lu Y, Zhu Z, Shi Y, Wang W. Distribution and Dietary Predictors of Urinary Phthalate Metabolites among Pregnant Women in Shanghai, China. Int J Environ Res Public Health. 2019;16(8). doi: 10.3390/ijerph16081366. PubMed PMID: 30995748; PMCID: PMC6518169. 92. Cantonwine DE, Cordero JF, Rivera-Gonzalez LO, Anzalota Del Toro LV, Ferguson KK, Mukherjee B, Calafat AM, Crespo N, Jimenez-Velez B, Padilla IY, Alshawabkeh AN, Meeker JD. Urinary phthalate metabolite concentrations among pregnant women in Northern Puerto Rico: distribution, temporal variability, and predictors. Environ Int. 2014;62:1-11. doi: 10.1016/j.envint.2013.09.014. PubMed PMID: 24161445; PMCID: PMC3874859. 93. Philips EM, Jaddoe VWV, Asimakopoulos AG, Kannan K, Steegers EAP, Santos S, Trasande L. Bisphenol and phthalate concentrations and its determinants among pregnant women in a population-based cohort in the Netherlands, 2004-5. Environ Res. 2018;161:562-72. Epub 2017/12/16. doi: 10.1016/j.envres.2017.11.051. PubMed PMID: 29245124; PMCID: PMC5820024. 94. Lee W-C, Fisher M, Davis K, Arbuckle TE, Sinha SK. Identification of chemical mixtures to which Canadian pregnant women are exposed: The MIREC Study. Environment International. 2017;99:321-30. doi: https://doi.org/10.1016/j.envint.2016.12.015. 95. Kalloo G, Wellenius GA, McCandless L, Calafat AM, Sjodin A, Karagas M, Chen A, Yolton K, Lanphear BP, Braun JM. Profiles and Predictors of Environmental Chemical Mixture Exposure among Pregnant Women: The Health Outcomes and Measures of the Environment Study. Environmental Science & Technology. 2018;52(17):10104-13. doi: 10.1021/acs.est.8b02946. 96. Montazeri P, Thomsen C, Casas M, de Bont J, Haug LS, Maitre L, Papadopoulou E, Sakhi AK, Slama R, Saulnier PJ, Urquiza J, Grazuleviciene R, Andrusaityte S, McEachan R, Wright J, Chatzi L, Basagana X, Vrijheid M. Socioeconomic position and exposure to multiple environmental chemical contaminants in six European mother-child 156 cohorts. Int J Hyg Environ Health. 2019;222(5):864-72. doi: 10.1016/j.ijheh.2019.04.002. PubMed PMID: 31010791. 97. Chen H, Zhang W, Zhou Y, Li J, Zhao H, Xu S, Xia W, Cai Z, Li Y. Characteristics of exposure to multiple environmental chemicals among pregnant women in Wuhan, China. Sci Total Environ. 2021;754:142167. Epub 2020/09/12. doi: 10.1016/j.scitotenv.2020.142167. PubMed PMID: 32916497. 98. Bannon AL, Waring ME, Leung K, Masiero JV, Stone JM, Scannell EC, Moore Simas TA. Comparison of Self-reported and Measured Pre-pregnancy Weight: Implications for Gestational Weight Gain Counseling. Matern Child Health J. 2017;21(7):1469-78. doi: 10.1007/s10995-017-2266-3. PubMed PMID: 28155023. 99. Holland E, Moore Simas TA, Doyle Curiale DK, Liao X, Waring ME. Self-reported pre-pregnancy weight versus weight measured at first prenatal visit: effects on categorization of pre-pregnancy body mass index. Matern Child Health J. 2013;17(10):1872-8. doi: 10.1007/s10995-012-1210-9. PubMed PMID: 23247668; PMCID: PMC3622142. 100. Natamba BK, Sanchez SE, Gelaye B, Williams MA. Concordance between self- reported pre-pregnancy body mass index (BMI) and BMI measured at the first prenatal study contact. BMC Pregnancy Childbirth. 2016;16(1):187. doi: 10.1186/s12884-016- 0983-z. PubMed PMID: 27460221; PMCID: PMC4962409. 101. Boucher B, Cotterchio M, Kreiger N, Nadalin V, Block T, Block G. Validity and reliability of the Block98 food-frequency questionnaire in a sample of Canadian women. Public Health Nutr. 2006;9(1):84-93. doi: 10.1079/phn2005763. PubMed PMID: 16480538. 102. McCullough ML, Feskanich D, Stampfer MJ, Giovannucci EL, Rimm EB, Hu FB, Spiegelman D, Hunter DJ, Colditz GA, Willett WC. Diet quality and major chronic disease risk in men and women: moving toward improved dietary guidance. Am J Clin Nutr. 2002;76(6):1261-71. doi: 10.1093/ajcn/76.6.1261. PubMed PMID: 12450892. 103. Chiuve SE, Fung TT, Rimm EB, Hu FB, McCullough ML, Wang M, Stampfer MJ, Willett WC. Alternative dietary indices both strongly predict risk of chronic disease. J Nutr. 2012;142(6):1009-18. Epub 2012/04/20. doi: 10.3945/jn.111.157222. PubMed PMID: 22513989; PMCID: PMC3738221. 104. Silva MJ, Samandar E, Preau JL, Jr., Reidy JA, Needham LL, Calafat AM. Quantification of 22 phthalate metabolites in human urine. J Chromatogr B Analyt Technol Biomed Life Sci. 2007;860(1):106-12. doi: 10.1016/j.jchromb.2007.10.023. PubMed PMID: 17997365. 105. Silva MJ, Jia T, Samandar E, Preau JL, Jr., Calafat AM. Environmental exposure to the plasticizer 1,2-cyclohexane dicarboxylic acid, diisononyl ester (DINCH) in U.S. adults (2000-2012). Environ Res. 2013;126:159-63. doi: 10.1016/j.envres.2013.05.007. PubMed PMID: 23777640; PMCID: PMC4554753. 157 106. Silva MJ, Wong LY, Samandar E, Preau JL, Jr., Jia LT, Calafat AM. Exposure to di-2-ethylhexyl terephthalate in the U.S. general population from the 2015-2016 National Health and Nutrition Examination Survey. Environ Int. 2019;123:141-7. doi: 10.1016/j.envint.2018.11.041. PubMed PMID: 30529838; PMCID: PMC7917578. 107. Succop PA, Clark S, Chen M, Galke W. Imputation of data values that are less than a detection limit. J Occup Environ Hyg. 2004;1(7):436-41. doi: 10.1080/15459620490462797. PubMed PMID: 15238313. 108. Weiss L, Arbuckle TE, Fisher M, Ramsay T, Mallick R, Hauser R, LeBlanc A, Walker M, Dumas P, Lang C. Temporal variability and sources of triclosan exposure in pregnancy. Int J Hyg Environ Health. 2015;218(6):507-13. Epub 2015/05/27. doi: 10.1016/j.ijheh.2015.04.003. PubMed PMID: 26009209. 109. Meeker JD, Hu H, Cantonwine DE, Lamadrid-Figueroa H, Calafat AM, Ettinger AS, Hernandez-Avila M, Loch-Caruso R, Tellez-Rojo MM. Urinary phthalate metabolites in relation to preterm birth in Mexico city. Environ Health Perspect. 2009;117(10):1587- 92. Epub 2009/12/19. doi: 10.1289/ehp.0800522. PubMed PMID: 20019910; PMCID: PMC2790514. 110. NHANES. National Health and Nutrition Examination Survey Data. In: (NCHS) NCfHS, editor. Hyattsville, Maryland: U.S. Department of Health and Human Services, Centers for Disease Control and Prevention; 2013-2014. 111. NHANES. National Health and Nutrition Examination Survey Data. In: (NCHS) NCfHS, editor. Hyattsville, Maryland: U.S. Department of Health and Human Services, Centers for Disease Control and Prevention; 2015-2016. 112. NHANES. National Health and Nutrition Examination Survey Data. In: (NCHS) NCfHS, editor. Hyattsville, Maryland: U.S. Department of Health and Human Services, Centers for Disease Control and Prevention; 2017-2018. 113. Rodriguez-Carmona Y, Ashrap P, Calafat AM, Ye X, Rosario Z, Bedrosian LD, Huerta-Montanez G, Velez-Vega CM, Alshawabkeh A, Cordero JF, Meeker JD, Watkins D. Determinants and characterization of exposure to phthalates, DEHTP and DINCH among pregnant women in the PROTECT birth cohort in Puerto Rico. J Expo Sci Environ Epidemiol. 2020;30(1):56-69. doi: 10.1038/s41370-019-0168-8. PubMed PMID: 31481681; PMCID: PMC6917904. 114. Gao H, Zhu YD, Xu YY, Zhang YW, Yao HY, Sheng J, Jin ZX, Ren LL, Huang K, Hao JH, Tao FB. Season-dependent concentrations of urinary phthalate metabolites among Chinese pregnant women: Repeated measures analysis. Environ Int. 2017;104:110-7. doi: 10.1016/j.envint.2017.03.021. PubMed PMID: 28389128. 115. Lyden GR, Barrett ES, Sathyanarayana S, Bush NR, Swan SH, Nguyen RHN. Pregnancy intention and phthalate metabolites among pregnant women in The Infant Development and Environment Study cohort. Paediatr Perinat Epidemiol. 158 2020;34(6):736-43. doi: 10.1111/ppe.12674. PubMed PMID: 32249967; PMCID: PMC7541656. 116. Weir CB, Jan A. BMI Classification Percentile And Cut Off Points. StatPearls. Treasure Island (FL)2021. 117. Wasserstein RL, Lazar NA. The ASA's Statement on p-Values: Context, Process, and Purpose. Am Stat. 2016;70(2):129-31. doi: 10.1080/00031305.2016.1154108. PubMed PMID: WOS:000378462300001. 118. Amrhein V, Greenland S, McShane B. Scientists rise up against statistical significance. Nature. 2019;567(7748):305-7. doi: 10.1038/d41586-019-00857-9. PubMed PMID: 30894741. 119. Polinski KJ, Dabelea D, Hamman RF, Adgate JL, Calafat AM, Ye X, Starling AP. Distribution and predictors of urinary concentrations of phthalate metabolites and phenols among pregnant women in the Healthy Start Study. Environ Res. 2018;162:308-17. doi: 10.1016/j.envres.2018.01.025. PubMed PMID: 29407762; PMCID: PMC5811372. 120. Lewin A, Arbuckle TE, Fisher M, Liang CL, Marro L, Davis K, Abdelouahab N, Fraser WD. Univariate predictors of maternal concentrations of environmental chemicals: The MIREC study. Int J Hyg Environ Health. 2017;220(2 Pt A):77-85. doi: 10.1016/j.ijheh.2017.01.001. PubMed PMID: 28109710. 121. Wu H, Kupsco AJ, Deierlein AL, Just AC, Calafat AM, Oken E, Braun JM, Mercado-Garcia A, Cantoral A, Tellez-Rojo MM, Wright RO, Baccarelli AA. Trends and Patterns of Phthalates and Phthalate Alternatives Exposure in Pregnant Women from Mexico City during 2007-2010. Environ Sci Technol. 2020;54(3):1740-9. doi: 10.1021/acs.est.9b05836. PubMed PMID: 31944681; PMCID: PMC7094762. 122. Smith AR, Kogut KR, Parra K, Bradman A, Holland N, Harley KG. Dietary intake and household exposures as predictors of urinary concentrations of high molecular weight phthalates and bisphenol A in a cohort of adolescents. J Expo Sci Environ Epidemiol. 2021. doi: 10.1038/s41370-021-00305-9. PubMed PMID: 33619365; PMCID: PMC8380263. 123. Yaghjyan L, Carlsson NP, Ghita GL, Chang SH. Associations of individual characteristics and lifestyle factors with metabolism of di-2-ethylhexyl phthalate in NHANES 2001-2012. Environ Res. 2016;149:23-31. doi: 10.1016/j.envres.2016.05.002. PubMed PMID: 27174780; PMCID: PMC5536839. 124. Arbuckle TE, Davis K, Marro L, Fisher M, Legrand M, LeBlanc A, Gaudreau E, Foster WG, Choeurng V, Fraser WD, Group MS. Phthalate and bisphenol A exposure among pregnant women in Canada--results from the MIREC study. Environ Int. 2014;68:55-65. doi: 10.1016/j.envint.2014.02.010. PubMed PMID: 24709781. 159 125. Zhu Y, Wan Y, Li Y, Zhang B, Zhou A, Cai Z, Qian Z, Zhang C, Huo W, Huang K, Hu J, Cheng L, Chang H, Huang Z, Xu B, Xia W, Xu S. Free and total urinary phthalate metabolite concentrations among pregnant women from the Healthy Baby Cohort (HBC), China. Environ Int. 2016;88:67-73. doi: 10.1016/j.envint.2015.12.004. PubMed PMID: 26722670. 126. Qu J, Xia W, Qian X, Wu Y, Li J, Wen S, Xu S. Geographic distribution and time trend of human exposure of Di(2-ethylhexyl) phthalate among different age groups based on global biomonitoring data. Chemosphere. 2022;287(Pt 2):132115. doi: 10.1016/j.chemosphere.2021.132115. PubMed PMID: 34826892. 127. Runkel AA, Mazej D, Snoj Tratnik J, Tkalec Z, Kosjek T, Horvat M. Exposure of men and lactating women to environmental phenols, phthalates, and DINCH. Chemosphere. 2022;286(Pt 3):131858. doi: 10.1016/j.chemosphere.2021.131858. PubMed PMID: 34399256. 128. Koch HM, Lorber M, Christensen KL, Palmke C, Koslitz S, Bruning T. Identifying sources of phthalate exposure with human biomonitoring: results of a 48h fasting study with urine collection and personal activity patterns. Int J Hyg Environ Health. 2013;216(6):672-81. Epub 2013/01/22. doi: 10.1016/j.ijheh.2012.12.002. PubMed PMID: 23333758. 129. Rosofsky A, Janulewicz P, Thayer KA, McClean M, Wise LA, Calafat AM, Mikkelsen EM, Taylor KW, Hatch EE. Exposure to multiple chemicals in a cohort of reproductive-aged Danish women. Environ Res. 2017;154:73-85. doi: 10.1016/j.envres.2016.12.011. PubMed PMID: 28039828; PMCID: PMC5328929. 130. Hsieh CJ, Chang YH, Hu A, Chen ML, Sun CW, Situmorang RF, Wu MT, Wang SL, group Ts. Personal care products use and phthalate exposure levels among pregnant women. Sci Total Environ. 2019;648:135-43. doi: 10.1016/j.scitotenv.2018.08.149. PubMed PMID: 30114584. 131. Parlett LE, Calafat AM, Swan SH. Women's exposure to phthalates in relation to use of personal care products. J Expo Sci Environ Epidemiol. 2013;23(2):197-206. doi: 10.1038/jes.2012.105. PubMed PMID: 23168567; PMCID: PMC4097177. 132. Campioli E, Lee S, Lau M, Marques L, Papadopoulos V. Effect of prenatal DINCH plasticizer exposure on rat offspring testicular function and metabolism. Scientific Reports. 2017;7(1):11072. doi: 10.1038/s41598-017-11325-7. 133. Woodward MJ, Obsekov V, Jacobson MH, Kahn LG, Trasande L. Phthalates and Sex Steroid Hormones Among Men From NHANES, 2013-2016. J Clin Endocrinol Metab. 2020;105(4). doi: 10.1210/clinem/dgaa039. PubMed PMID: 31996892; PMCID: PMC7067547. 134. Rodríguez-Carmona Y, Ashrap P, Calafat AM, Ye X, Rosario Z, Bedrosian LD, Huerta-Montanez G, Vélez-Vega CM, Alshawabkeh A, Cordero JF, Meeker JD, Watkins D. Determinants and characterization of exposure to phthalates, DEHTP and DINCH 160 among pregnant women in the PROTECT birth cohort in Puerto Rico. Journal of Exposure Science & Environmental Epidemiology. 2020;30(1):56-69. doi: 10.1038/s41370-019-0168-8. 135. Dickinson A, MacKay D. Health habits and other characteristics of dietary supplement users: a review. Nutr J. 2014;13:14. Epub 2014/02/07. doi: 10.1186/1475- 2891-13-14. PubMed PMID: 24499096; PMCID: PMC3931917. 136. Kelley KE, Hernandez-Diaz S, Chaplin EL, Hauser R, Mitchell AA. Identification of phthalates in medications and dietary supplement formulations in the United States and Canada. Environ Health Perspect. 2012;120(3):379-84. Epub 2011/12/16. doi: 10.1289/ehp.1103998. PubMed PMID: 22169271; PMCID: PMC3295354. 137. Yan X, Calafat A, Lashley S, Smulian J, Ananth C, Barr D, Silva M, Ledoux T, Hore P, Robson MG. Phthalates Biomarker Identification and Exposure Estimates in a Population of Pregnant Women. Hum Ecol Risk Assess. 2009;15(3):565-78. Epub 2009/05/01. doi: 10.1080/10807030902892554. PubMed PMID: 20686649; PMCID: PMC2913903. 138. Boss J, Zhai J, Aung MT, Ferguson KK, Johns LE, McElrath TF, Meeker JD, Mukherjee B. Associations between mixtures of urinary phthalate metabolites with gestational age at delivery: a time to event analysis using summative phthalate risk scores. Environ Health. 2018;17(1):56. doi: 10.1186/s12940-018-0400-3. PubMed PMID: 29925380; PMCID: PMC6011420. 139. Fromme H, Bolte G, Koch HM, Angerer J, Boehmer S, Drexler H, Mayer R, Liebl B. Occurrence and daily variation of phthalate metabolites in the urine of an adult population. International Journal of Hygiene and Environmental Health. 2007;210(1):21- 33. doi: https://doi.org/10.1016/j.ijheh.2006.09.005. 140. Johns LE, Cooper GS, Galizia A, Meeker JD. Exposure assessment issues in epidemiology studies of phthalates. Environment international. 2015;85:27-39. doi: 10.1016/j.envint.2015.08.005. PubMed PMID: 26313703. 141. Preau JL, Jr., Wong LY, Silva MJ, Needham LL, Calafat AM. Variability over 1 week in the urinary concentrations of metabolites of diethyl phthalate and di(2- ethylhexyl) phthalate among eight adults: an observational study. Environ Health Perspect. 2010;118(12):1748-54. Epub 2010/08/28. doi: 10.1289/ehp.1002231. PubMed PMID: 20797930; PMCID: PMC3002195. 142. Noyola-Martinez N, Halhali A, Barrera D. Steroid hormones and pregnancy. Gynecol Endocrinol. 2019;35(5):376-84. doi: 10.1080/09513590.2018.1564742. PubMed PMID: 30793997. 143. Xu X, Veenstra TD, Fox SD, Roman JM, Issaq HJ, Falk R, Saavedra JE, Keefer LK, Ziegler RG. Measuring fifteen endogenous estrogens simultaneously in human urine by high-performance liquid chromatography-mass spectrometry. Anal Chem. 2005;77(20):6646-54. doi: 10.1021/ac050697c. PubMed PMID: 16223252. 161 144. O'Leary P, Boyne P, Flett P, Beilby J, James I. Longitudinal assessment of changes in reproductive hormones during normal pregnancy. Clin Chem. 1991;37(5):667-72. PubMed PMID: 1827758. 145. (CDC) CfDCaP. National Health and Nutrition Examination Survey Data. In: (NCHS) NCfHS, editor. Hyattsville, Maryland: U.S. Department of Health and Human Services, Centers for Disease Control and Prevention; 2015-2016. 146. (CDC) CfDCaP. National Report on Human Exposure to Environmental Chemicals, Updated Tables January 2019. Available at https://www.cdc.gov/exposurereport/index.html. 147. Benjamin S, Masai E, Kamimura N, Takahashi K, Anderson RC, Faisal PA. Phthalates impact human health: Epidemiological evidences and plausible mechanism of action. J Hazard Mater. 2017;340:360-83. doi: 10.1016/j.jhazmat.2017.06.036. PubMed PMID: 28800814. 148. Wang W, Craig ZR, Basavarajappa MS, Gupta RK, Flaws JA. Di (2-ethylhexyl) phthalate inhibits growth of mouse ovarian antral follicles through an oxidative stress pathway. Toxicol Appl Pharmacol. 2012;258(2):288-95. doi: 10.1016/j.taap.2011.11.008. PubMed PMID: 22155089; PMCID: PMC3259146. 149. Marsee K, Woodruff TJ, Axelrad DA, Calafat AM, Swan SH. Estimated daily phthalate exposures in a population of mothers of male infants exhibiting reduced anogenital distance. Environ Health Perspect. 2006;114(6):805-9. doi: 10.1289/ehp.8663. PubMed PMID: 16759976; PMCID: PMC1480516. 150. Sathyanarayana S, Butts S, Wang C, Barrett E, Nguyen R, Schwartz SM, Haaland W, Swan SH, Team T. Early Prenatal Phthalate Exposure, Sex Steroid Hormones, and Birth Outcomes. J Clin Endocrinol Metab. 2017;102(6):1870-8. Epub 2017/03/23. doi: 10.1210/jc.2016-3837. PubMed PMID: 28324030; PMCID: PMC5470772. 151. Sathyanarayana S, Barrett E, Butts S, Wang C, Swan SH. Phthalate exposure and reproductive hormone concentrations in pregnancy. Reproduction. 2014;147(4):401-9. Epub 2013/11/08. doi: 10.1530/REP-13-0415. PubMed PMID: 24196015; PMCID: PMC3943643. 152. Cathey AL, Watkins D, Rosario ZY, Velez C, Alshawabkeh AN, Cordero JF, Meeker JD. Associations of Phthalates and Phthalate Replacements With CRH and Other Hormones Among Pregnant Women in Puerto Rico. J Endocr Soc. 2019;3(6):1127-49. doi: 10.1210/js.2019-00010. PubMed PMID: 31093596; PMCID: PMC6510018. 153. Zota AR, Calafat AM, Woodruff TJ. Temporal trends in phthalate exposures: findings from the National Health and Nutrition Examination Survey, 2001-2010. Environ Health Perspect. 2014;122(3):235-41. doi: 10.1289/ehp.1306681. PubMed PMID: 24425099; PMCID: PMC3948032. 162 154. Commission USCPS. Prohibition of Children’s Toys and Child Care Articles Containing Specified Phthalates: Determinations Regarding Certain Plastics. Rules and Regulations ed2017. p. 41163-72. 155. Banker M, Puttabyatappa M, O'Day P, Goodrich JM, Kelley AS, Domino SE, Smith YR, Dolinoy DC, Song PXK, Auchus RJ, Padmanabhan V. Association of Maternal-Neonatal Steroids with Early Pregnancy Endocrine Disrupting Chemicals and Pregnancy Outcomes. J Clin Endocrinol Metab. 2020. doi: 10.1210/clinem/dgaa909. PubMed PMID: 33280001. 156. Coburn SB, Stanczyk FZ, Falk RT, McGlynn KA, Brinton LA, Sampson J, Bradwin G, Xu X, Trabert B. Comparability of serum, plasma, and urinary estrogen and estrogen metabolite measurements by sex and menopausal status. Cancer Causes Control. 2019;30(1):75-86. doi: 10.1007/s10552-018-1105-1. PubMed PMID: 30506492; PMCID: PMC6447065. 157. Fortner RT, Hankinson SE, Schairer C, Xu X, Ziegler RG, Eliassen AH. Association between reproductive factors and urinary estrogens and estrogen metabolites in premenopausal women. Cancer Epidemiol Biomarkers Prev. 2012;21(6):959-68. doi: 10.1158/1055-9965.EPI-12-0171. PubMed PMID: 22454378; PMCID: PMC3381957. 158. Gu F, Caporaso NE, Schairer C, Fortner RT, Xu X, Hankinson SE, Eliassen AH, Ziegler RG. Urinary concentrations of estrogens and estrogen metabolites and smoking in caucasian women. Cancer Epidemiol Biomarkers Prev. 2013;22(1):58-68. doi: 10.1158/1055-9965.EPI-12-0909. PubMed PMID: 23104668; PMCID: PMC3643002. 159. Hall Moran V, Leathard HL, Coley J. Urinary hormone levels during the natural menstrual cycle: the effect of age. J Endocrinol. 2001;170(1):157-64. doi: 10.1677/joe.0.1700157. PubMed PMID: 11431148. 160. Cantonwine DE, McElrath TF, Trabert B, Xu X, Sampson J, Roberts JM, Hoover RN, Troisi R. Estrogen metabolism pathways in preeclampsia and normal pregnancy. Steroids. 2019;144:8-14. doi: 10.1016/j.steroids.2019.01.005. PubMed PMID: 30685337; PMCID: PMC6681456. 161. Eliassen AH, Spiegelman D, Xu X, Keefer LK, Veenstra TD, Barbieri RL, Willett WC, Hankinson SE, Ziegler RG. Urinary estrogens and estrogen metabolites and subsequent risk of breast cancer among premenopausal women. Cancer Res. 2012;72(3):696-706. doi: 10.1158/0008-5472.CAN-11-2507. PubMed PMID: 22144471; PMCID: PMC3271178. 162. Shin HM, Bennett DH, Barkoski J, Ye X, Calafat AM, Tancredi D, Hertz-Picciotto I. Variability of urinary concentrations of phthalate metabolites during pregnancy in first morning voids and pooled samples. Environ Int. 2019;122:222-30. doi: 10.1016/j.envint.2018.11.012. PubMed PMID: 30477814; PMCID: PMC6311426. 163 163. Wesselink AK, Fruh V, Hauser R, Weuve J, Taylor KW, Orta OR, Claus Henn B, Bethea TN, McClean MD, Williams PL, Calafat AM, Baird DD, Wise LA. Correlates of urinary concentrations of phthalate and phthalate alternative metabolites among reproductive-aged Black women from Detroit, Michigan. J Expo Sci Environ Epidemiol. 2020. doi: 10.1038/s41370-020-00270-9. PubMed PMID: 32980856. 164. Mistry HD, Eisele N, Escher G, Dick B, Surbek D, Delles C, Currie G, Schlembach D, Mohaupt MG, Gennari-Moser C. Gestation-specific reference intervals for comprehensive spot urinary steroid hormone metabolite analysis in normal singleton pregnancy and 6 weeks postpartum. Reprod Biol Endocrinol. 2015;13:101. doi: 10.1186/s12958-015-0100-6. PubMed PMID: 26337185; PMCID: PMC4559160. 165. Villarroel C, Salinas A, Lopez P, Kohen P, Rencoret G, Devoto L, Codner E. Pregestational type 2 diabetes and gestational diabetes exhibit different sexual steroid profiles during pregnancy. Gynecol Endocrinol. 2017;33(3):212-7. Epub 2016/11/30. doi: 10.1080/09513590.2016.1248933. PubMed PMID: 27898283. 166. Rothman KJ. No adjustments are needed for multiple comparisons. Epidemiology. 1990;1(1):43-6. PubMed PMID: 2081237. 167. Ziegler RG, Faupel-Badger JM, Sue LY, Fuhrman BJ, Falk RT, Boyd-Morin J, Henderson MK, Hoover RN, Veenstra TD, Keefer LK, Xu X. A new approach to measuring estrogen exposure and metabolism in epidemiologic studies. J Steroid Biochem Mol Biol. 2010;121(3-5):538-45. doi: 10.1016/j.jsbmb.2010.03.068. PubMed PMID: 20382222; PMCID: PMC6276800. 168. Cohn BA, Cirillo PM, Hopper BR, Siiteri PK. Third Trimester Estrogens and Maternal Breast Cancer: Prospective Evidence. J Clin Endocrinol Metab. 2017;102(10):3739-48. doi: 10.1210/jc.2016-3476. PubMed PMID: 28973345; PMCID: PMC5630249. 169. Saadeldin IM, Hussein MA, Suleiman AH, Abohassan MG, Ahmed MM, Moustafa AA, Moumen AF, Abdel-Aziz Swelum A. Ameliorative effect of ginseng extract on phthalate and bisphenol A reprotoxicity during pregnancy in rats. Environ Sci Pollut Res Int. 2018;25(21):21205-15. doi: 10.1007/s11356-018-2299-1. PubMed PMID: 29777495. 170. Kragie L. Aromatase in primate pregnancy: a review. Endocr Res. 2002;28(3):121-8. doi: 10.1081/erc-120015041. PubMed PMID: 12489562. 171. Perez-Albaladejo E, Fernandes D, Lacorte S, Porte C. Comparative toxicity, oxidative stress and endocrine disruption potential of plasticizers in JEG-3 human placental cells. Toxicol In Vitro. 2017;38:41-8. doi: 10.1016/j.tiv.2016.11.003. PubMed PMID: 27825933. 172. Engel A, Buhrke T, Kasper S, Behr AC, Braeuning A, Jessel S, Seidel A, Volkel W, Lampen A. The urinary metabolites of DINCH((R)) have an impact on the activities of the human nuclear receptors ERalpha, ERbeta, AR, PPARalpha and PPARgamma. 164 Toxicol Lett. 2018;287:83-91. doi: 10.1016/j.toxlet.2018.02.006. PubMed PMID: 29421333. 173. Strakovsky RS, Schantz SL. Using Experimental Models to Assess Effects of Bisphenol A (BPA) and Phthalates on the Placenta: Challenges and Perspectives. Toxicol Sci. 2018;166(2):250-68. doi: 10.1093/toxsci/kfy224. PubMed PMID: 30203063; PMCID: PMC6260157. 174. Qian Y, Shao H, Ying X, Huang W, Hua Y. The Endocrine Disruption of Prenatal Phthalate Exposure in Mother and Offspring. Front Public Health. 2020;8:366. doi: 10.3389/fpubh.2020.00366. PubMed PMID: 32984231; PMCID: PMC7483495. 175. Vernet C, Philippat C, Agier L, Calafat AM, Ye X, Lyon-Caen S, Hainaut P, Siroux V, Schisterman EF, Slama R. An Empirical Validation of the Within-subject Biospecimens Pooling Approach to Minimize Exposure Misclassification in Biomarker- based Studies. Epidemiology. 2019;30(5):756-67. doi: 10.1097/EDE.0000000000001056. PubMed PMID: 31373935. 176. MacPherson S, Arbuckle TE, Fisher M. Adjusting urinary chemical biomarkers for hydration status during pregnancy. J Expo Sci Environ Epidemiol. 2018;28(5):481-93. Epub 2018/06/09. doi: 10.1038/s41370-018-0043-z. PubMed PMID: 29880833. 177. Margerison Zilko CE, Rehkopf D, Abrams B. Association of maternal gestational weight gain with short- and long-term maternal and child health outcomes. Am J Obstet Gynecol. 2010;202(6):574 e1-8. doi: 10.1016/j.ajog.2009.12.007. PubMed PMID: 20132923. 178. Hutchins F, Abrams B, Brooks M, Colvin A, Moore Simas T, Rosal M, Sternfeld B, Crawford S. The Effect of Gestational Weight Gain Across Reproductive History on Maternal Body Mass Index in Midlife: The Study of Women's Health Across the Nation. J Womens Health (Larchmt). 2020;29(2):148-57. doi: 10.1089/jwh.2019.7839. PubMed PMID: 31794347; PMCID: PMC7045562. 179. Soomro MH, Maesano CN, Heude B, Bornehag CG, Annesi-Maesano I. The association between maternal urinary phthalate metabolites concentrations and pregnancy induced hypertension: Results from the EDEN Mother-Child Cohort. J Gynecol Obstet Hum Reprod. 2021;50(10):102216. Epub 2021/09/06. doi: 10.1016/j.jogoh.2021.102216. PubMed PMID: 34482002. 180. Werner EF, Braun JM, Yolton K, Khoury JC, Lanphear BP. The association between maternal urinary phthalate concentrations and blood pressure in pregnancy: The HOME Study. Environ Health. 2015;14:75. Epub 2015/09/19. doi: 10.1186/s12940- 015-0062-3. PubMed PMID: 26380974; PMCID: PMC4574131. 181. Shaffer RM, Ferguson KK, Sheppard L, James-Todd T, Butts S, Chandrasekaran S, Swan SH, Barrett ES, Nguyen R, Bush N, McElrath TF, Sathyanarayana S, team TS. Maternal urinary phthalate metabolites in relation to gestational diabetes and glucose 165 intolerance during pregnancy. Environ Int. 2019;123:588-96. Epub 2019/01/10. doi: 10.1016/j.envint.2018.12.021. PubMed PMID: 30622083; PMCID: PMC6347428. 182. James-Todd TM, Chiu YH, Messerlian C, Minguez-Alarcon L, Ford JB, Keller M, Petrozza J, Williams PL, Ye X, Calafat AM, Hauser R, Team ES. Trimester-specific phthalate concentrations and glucose levels among women from a fertility clinic. Environ Health. 2018;17(1):55. Epub 2018/06/15. doi: 10.1186/s12940-018-0399-5. PubMed PMID: 29898728; PMCID: PMC6000948. 183. Fisher BG, Frederiksen H, Andersson AM, Juul A, Thankamony A, Ong KK, Dunger DB, Hughes IA, Acerini CL. Serum Phthalate and Triclosan Levels Have Opposing Associations With Risk Factors for Gestational Diabetes Mellitus. Front Endocrinol (Lausanne). 2018;9:99. Epub 2018/03/30. doi: 10.3389/fendo.2018.00099. PubMed PMID: 29593656; PMCID: PMC5859030. 184. Bui TT, Giovanoulis G, Cousins AP, Magner J, Cousins IT, de Wit CA. Human exposure, hazard and risk of alternative plasticizers to phthalate esters. Sci Total Environ. 2016;541:451-67. doi: 10.1016/j.scitotenv.2015.09.036. PubMed PMID: 26410720. 185. Lemke N, Murawski A, Lange R, Weber T, Apel P, Debiak M, Koch HM, Kolossa- Gehring M. Substitutes mimic the exposure behaviour of REACH regulated phthalates - A review of the German HBM system on the example of plasticizers. Int J Hyg Environ Health. 2021;236:113780. Epub 2021/06/15. doi: 10.1016/j.ijheh.2021.113780. PubMed PMID: 34126298. 186. Zhang Y, Mustieles V, Yland J, Braun JM, Williams PL, Attaman JA, Ford JB, Calafat AM, Hauser R, Messerlian C. Association of Parental Preconception Exposure to Phthalates and Phthalate Substitutes With Preterm Birth. JAMA Netw Open. 2020;3(4):e202159. doi: 10.1001/jamanetworkopen.2020.2159. PubMed PMID: 32259265; PMCID: PMC7139277. 187. Fruh V, Claus Henn B, Weuve J, Wesselink AK, Orta OR, Heeren T, Hauser R, Calafat AM, Williams PL, Baird DD, Wise LA. Incidence of uterine leiomyoma in relation to urinary concentrations of phthalate and phthalate alternative biomarkers: A prospective ultrasound study. Environ Int. 2021;147:106218. doi: 10.1016/j.envint.2020.106218. PubMed PMID: 33360166. 188. Machtinger R, Gaskins AJ, Racowsky C, Mansur A, Adir M, Baccarelli AA, Calafat AM, Hauser R. Urinary concentrations of biomarkers of phthalates and phthalate alternatives and IVF outcomes. Environ Int. 2018;111:23-31. doi: 10.1016/j.envint.2017.11.011. PubMed PMID: 29161633; PMCID: PMC5800972. 189. Gore AC, Chappell VA, Fenton SE, Flaws JA, Nadal A, Prins GS, Toppari J, Zoeller RT. Executive Summary to EDC-2: The Endocrine Society's Second Scientific Statement on Endocrine-Disrupting Chemicals. Endocr Rev. 2015;36(6):593-602. Epub 2015/09/29. doi: 10.1210/er.2015-1093. PubMed PMID: 26414233; PMCID: PMC4702495. 166 190. Maradonna F, Carnevali O. Lipid Metabolism Alteration by Endocrine Disruptors in Animal Models: An Overview. Front Endocrinol (Lausanne). 2018;9:654. Epub 2018/11/24. doi: 10.3389/fendo.2018.00654. PubMed PMID: 30467492; PMCID: PMC6236061. 191. Mohammadi H, Ashari S. Mechanistic insight into toxicity of phthalates, the involved receptors, and the role of Nrf2, NF-kappaB, and PI3K/AKT signaling pathways. Environ Sci Pollut Res Int. 2021;28(27):35488-527. Epub 2021/05/24. doi: 10.1007/s11356-021-14466-5. PubMed PMID: 34024001. 192. Philips EM, Santos S, Steegers EAP, Asimakopoulos AG, Kannan K, Trasande L, Jaddoe VWV. Maternal bisphenol and phthalate urine concentrations and weight gain during pregnancy. Environment International. 2020;135:105342. doi: https://doi.org/10.1016/j.envint.2019.105342. 193. Bellavia A, Hauser R, Seely EW, Meeker JD, Ferguson KK, McElrath TF, James- Todd T. Urinary phthalate metabolite concentrations and maternal weight during early pregnancy. Int J Hyg Environ Health. 2017;220(8):1347-55. Epub 2017/09/25. doi: 10.1016/j.ijheh.2017.09.005. PubMed PMID: 28939183; PMCID: PMC5701657. 194. Li J, Qian X, Zhao H, Zhou Y, Xu S, Li Y, Xiang L, Shi J, Xia W, Cai Z. Determinants of exposure levels, metabolism, and health risks of phthalates among pregnant women in Wuhan, China. Ecotoxicology and environmental safety. 2019;184:109657. Epub 2019/09/19. doi: 10.1016/j.ecoenv.2019.109657. PubMed PMID: 31526923. 195. Zukin H, Eskenazi B, Holland N, Harley KG. Prenatal exposure to phthalates and maternal metabolic outcomes in a high-risk pregnant Latina population. Environ Res. 2021;194:110712. Epub 2021/01/19. doi: 10.1016/j.envres.2021.110712. PubMed PMID: 33460632; PMCID: PMC7946766. 196. Tyagi P, James-Todd T, Minguez-Alarcon L, Ford JB, Keller M, Petrozza J, Calafat AM, Hauser R, Williams PL, Bellavia A, team Es. Identifying windows of susceptibility to endocrine disrupting chemicals in relation to gestational weight gain among pregnant women attending a fertility clinic. Environ Res. 2021;194:110638. Epub 2020/12/29. doi: 10.1016/j.envres.2020.110638. PubMed PMID: 33359703; PMCID: PMC7946748. 197. Bobb JF, Valeri L, Claus Henn B, Christiani DC, Wright RO, Mazumdar M, Godleski JJ, Coull BA. Bayesian kernel machine regression for estimating the health effects of multi-pollutant mixtures. Biostatistics. 2015;16(3):493-508. doi: 10.1093/biostatistics/kxu058. PubMed PMID: 25532525; PMCID: PMC5963470. 198. Warrington NM, Richmond R, Fenstra B, Myhre R, Gaillard R, Paternoster L, Wang CA, Beaumont RN, Das S, Murcia M, Barton SJ, Espinosa A, Thiering E, Atalay M, Pitkanen N, Ntalla I, Jonsson AE, Freathy R, Karhunen V, Tiesler CMT, Allard C, Crawford A, Ring SM, Melbye M, Magnus P, Rivadeneira F, Skotte L, Hansen T, Marsh J, Guxens M, Holloway JW, Grallert H, Jaddoe VWV, Lowe WL, Jr., Roumeliotaki T, 167 Hattersley AT, Lindi V, Pahkala K, Panoutsopoulou K, Standl M, Flexeder C, Bouchard L, Aagaard Nohr E, Marina LS, Kogevinas M, Niinikoski H, Dedoussis G, Heinrich J, Reynolds RM, Lakka T, Zeggini E, Raitakari OT, Chatzi L, Inskip HM, Bustamante M, Hivert MF, Jarvelin MR, Sorensen TIA, Pennell C, Felix JF, Jacobsson B, Geller F, Evans DM, Lawlor DA. Maternal and fetal genetic contribution to gestational weight gain. Int J Obes (Lond). 2018;42(4):775-84. doi: 10.1038/ijo.2017.248. PubMed PMID: 28990592; PMCID: PMC5784805. 199. Mando C, Calabrese S, Mazzocco MI, Novielli C, Anelli GM, Antonazzo P, Cetin I. Sex specific adaptations in placental biometry of overweight and obese women. Placenta. 2016;38:1-7. doi: 10.1016/j.placenta.2015.12.008. PubMed PMID: 26907375. 200. Grunebaum A, Dudenhausen J, Skupski DW. Impact of fetal gender on maternal weight gain during pregnancy. Am J Obstet Gynecol. 2014;210(1):S118-S. doi: 10.1016/j.ajog.2013.10.254. PubMed PMID: WOS:000330322600223. 201. Springel EH, Berggren EK, Huston-Presley L, Catalano PM. What is the impact of fetal sex on maternal glucose metabolism? Am J Obstet Gynecol. 2016;214(1):S307- S. doi: 10.1016/j.ajog.2015.10.615. PubMed PMID: WOS:000367092800564. 202. Pacyga DC, Haggerty DK, Nicol M, Henning M, Calafat AM, Braun JM, Schantz SL, Strakovsky RS. Identification of profiles and determinants of maternal pregnancy urinary biomarkers of phthalates and replacements in the Illinois Kids Development Study. Environ Int. 2022;162:107150. doi: 10.1016/j.envint.2022.107150. PubMed PMID: 35247685; PMCID: PMC8967784. 203. Santos S, Eekhout I, Voerman E, Gaillard R, Barros H, Charles MA, Chatzi L, Chevrier C, Chrousos GP, Corpeleijn E, Costet N, Crozier S, Doyon M, Eggesbo M, Fantini MP, Farchi S, Forastiere F, Gagliardi L, Georgiu V, Godfrey KM, Gori D, Grote V, Hanke W, Hertz-Picciotto I, Heude B, Hivert MF, Hryhorczuk D, Huang RC, Inskip H, Jusko TA, Karvonen AM, Koletzko B, Kupers LK, Lagstrom H, Lawlor DA, Lehmann I, Lopez-Espinosa MJ, Magnus P, Majewska R, Makela J, Manios Y, McDonald SW, Mommers M, Morgen CS, Moschonis G, Murinova L, Newnham J, Nohr EA, Andersen AN, Oken E, Oostvogels A, Pac A, Papadopoulou E, Pekkanen J, Pizzi C, Polanska K, Porta D, Richiardi L, Rifas-Shiman SL, Roeleveld N, Santa-Marina L, Santos AC, Smit HA, Sorensen TIA, Standl M, Stanislawski M, Stoltenberg C, Thiering E, Thijs C, Torrent M, Tough SC, Trnovec T, van Gelder M, van Rossem L, von Berg A, Vrijheid M, Vrijkotte TGM, Zvinchuk O, van Buuren S, Jaddoe VWV. Gestational weight gain charts for different body mass index groups for women in Europe, North America, and Oceania. BMC Med. 2018;16(1):201. Epub 2018/11/07. doi: 10.1186/s12916-018-1189- 1. PubMed PMID: 30396358; PMCID: PMC6217770. 204. Cohen S, Kamarck T, Mermelstein R. A Global Measure of Perceived Stress. J Health Soc Behav. 1983;24(4):385-96. doi: Doi 10.2307/2136404. PubMed PMID: WOS:A1983RZ58200008. 168 205. Cohen S, Williamson GM. Perceived Stress in a Probability Sample of the United-States. Clar Symp. 1988:31-67. PubMed PMID: WOS:A1988BQ68A00003. 206. Laraia BA, Bodnar LM, Siega-Riz AM. Pregravid body mass index is negatively associated with diet quality during pregnancy. Public Health Nutr. 2007;10(9):920-6. doi: 10.1017/S1368980007657991. PubMed PMID: 17381955. 207. Bodnar LM, Siega-Riz AM. A Diet Quality Index for Pregnancy detects variation in diet and differences by sociodemographic factors. Public Health Nutr. 2002;5(6):801- 9. doi: 10.1079/PHN2002348. PubMed PMID: 12570888. 208. VanderWeele TJ, Robins JM. Signed directed acyclic graphs for causal inference. J R Stat Soc Series B Stat Methodol. 2010;72(1):111-27. Epub 2010/01/01. doi: 10.1111/j.1467-9868.2009.00728.x. PubMed PMID: 25419168; PMCID: PMC4239133. 209. Tanner EM, Bornehag CG, Gennings C. Repeated holdout validation for weighted quantile sum regression. MethodsX. 2019;6:2855-60. Epub 2019/12/25. doi: 10.1016/j.mex.2019.11.008. PubMed PMID: 31871919; PMCID: PMC6911906. 210. Czarnota J, Gennings C, Colt JS, De Roos AJ, Cerhan JR, Severson RK, Hartge P, Ward MH, Wheeler DC. Analysis of Environmental Chemical Mixtures and Non- Hodgkin Lymphoma Risk in the NCI-SEER NHL Study. Environ Health Perspect. 2015;123(10):965-70. Epub 2015/03/10. doi: 10.1289/ehp.1408630. PubMed PMID: 25748701; PMCID: PMC4590749. 211. Hutcheon JA, Bodnar LM. Good Practices for Observational Studies of Maternal Weight and Weight Gain in Pregnancy. Paediatr Perinat Epidemiol. 2018;32(2):152-60. doi: 10.1111/ppe.12439. PubMed PMID: 29345321; PMCID: PMC5902633. 212. Gaston SA, Birnbaum LS, Jackson CL. Synthetic Chemicals and Cardiometabolic Health Across the Life Course Among Vulnerable Populations: a Review of the Literature from 2018 to 2019. Curr Environ Health Rep. 2020;7(1):30-47. doi: 10.1007/s40572-020-00265-6. PubMed PMID: 32037478; PMCID: PMC7187897. 213. Muscogiuri G, Barrea L, Laudisio D, Savastano S, Colao A. Obesogenic endocrine disruptors and obesity: myths and truths. Arch Toxicol. 2017;91(11):3469-75. doi: 10.1007/s00204-017-2071-1. PubMed PMID: 28975368. 214. Strakovsky RS, Schantz SL. Impacts of bisphenol A (BPA) and phthalate exposures on epigenetic outcomes in the human placenta. Environ Epigenet. 2018;4(3):dvy022. doi: 10.1093/eep/dvy022. PubMed PMID: 30210810; PMCID: PMC6128378. 215. Welch BM, Keil AP, Bommarito PA, van T' Erve TJ, Deterding LJ, Williams JG, Lih FB, Cantonwine DE, McElrath TF, Ferguson KK. Longitudinal exposure to consumer product chemicals and changes in plasma oxylipins in pregnant women. Environ Int. 169 2021;157:106787. doi: 10.1016/j.envint.2021.106787. PubMed PMID: 34314981; PMCID: PMC8490329. 216. Wang JQ, Hu YB, Gao H, Sheng J, Huang K, Zhang YW, Mao LJ, Zhou SS, Cai XX, Zhang LJ, Wang SF, Hao JH, Yang LQ, Tao FB. Sex-specific difference in placental inflammatory transcriptional biomarkers of maternal phthalate exposure: a prospective cohort study. J Expo Sci Environ Epidemiol. 2020;30(5):835-44. doi: 10.1038/s41370- 020-0200-z. PubMed PMID: 32015430. 217. Gao H, Tong J, Zhu BB, Geng ML, Gan H, Sun L, Wu XY, Huang K, Cao H, Liu WW, Tao SM, Ding P, Zhu P, Hao JH, Tao FB. Sex-specific mediation of placental inflammatory biomarkers in the effects of prenatal phthalate coexposure on preschooler cognitive development. Environ Sci Pollut Res Int. 2022;29(9):13305-14. doi: 10.1007/s11356-021-16695-0. PubMed PMID: 34585354. 218. Hlisnikova H, Petrovicova I, Kolena B, Sidlovska M, Sirotkin A. Effects and Mechanisms of Phthalates' Action on Reproductive Processes and Reproductive Health: A Literature Review. Int J Environ Res Public Health. 2020;17(18). Epub 2020/09/24. doi: 10.3390/ijerph17186811. PubMed PMID: 32961939; PMCID: PMC7559247. 219. Goldstein RF, Abell SK, Ranasinha S, Misso ML, Boyle JA, Harrison CL, Black MH, Li N, Hu G, Corrado F, Hegaard H, Kim YJ, Haugen M, Song WO, Kim MH, Bogaerts A, Devlieger R, Chung JH, Teede HJ. Gestational weight gain across continents and ethnicity: systematic review and meta-analysis of maternal and infant outcomes in more than one million women. BMC Med. 2018;16(1):153. doi: 10.1186/s12916-018-1128-1. PubMed PMID: 30165842; PMCID: PMC6117916. 220. Hack M, Klein NK, Taylor HG. Long-term developmental outcomes of low birth weight infants. Future Child. 1995;5(1):176-96. Epub 1995/01/01. PubMed PMID: 7543353. 221. Palatianou ME, Simos YV, Andronikou SK, Kiortsis DN. Long-term metabolic effects of high birth weight: a critical review of the literature. Horm Metab Res. 2014;46(13):911-20. doi: 10.1055/s-0034-1395561. PubMed PMID: 25473824. 222. Torniainen M, Wegelius A, Tuulio-Henriksson A, Lonnqvist J, Suvisaari J. Both low birthweight and high birthweight are associated with cognitive impairment in persons with schizophrenia and their first-degree relatives. Psychol Med. 2013;43(11):2361-7. doi: 10.1017/S0033291713000032. PubMed PMID: 23360614. 223. Gallacher DJ, Hart K, Kotecha S. Common respiratory conditions of the newborn. Breathe (Sheff). 2016;12(1):30-42. doi: 10.1183/20734735.000716. PubMed PMID: 27064402; PMCID: PMC4818233. 224. Bodnar LM, Cartus AR, Parisi SM, Abrams B, Himes KP, Eckhardt CL, Braxter B, Hutcheon JA. Pregnancy weight gain in twin gestations and maternal and child health 170 outcomes at 5 years. Int J Obes (Lond). 2021;45(7):1382-91. Epub 2021/03/05. doi: 10.1038/s41366-021-00792-8. PubMed PMID: 33658683; PMCID: PMC8238784. 225. Pugh SJ, Richardson GA, Hutcheon JA, Himes KP, Brooks MM, Day NL, Bodnar LM. Maternal Obesity and Excessive Gestational Weight Gain Are Associated with Components of Child Cognition. J Nutr. 2015;145(11):2562-9. Epub 2015/10/02. doi: 10.3945/jn.115.215525. PubMed PMID: 26423736; PMCID: PMC4620725. 226. Vandenberg LN, Colborn T, Hayes TB, Heindel JJ, Jacobs DR, Jr., Lee DH, Shioda T, Soto AM, vom Saal FS, Welshons WV, Zoeller RT, Myers JP. Hormones and endocrine-disrupting chemicals: low-dose effects and nonmonotonic dose responses. Endocr Rev. 2012;33(3):378-455. Epub 2012/03/16. doi: 10.1210/er.2011-1050. PubMed PMID: 22419778; PMCID: PMC3365860. 227. Paterni I, Granchi C, Katzenellenbogen JA, Minutolo F. Estrogen receptors alpha (ERalpha) and beta (ERbeta): subtype-selective ligands and clinical potential. Steroids. 2014;90:13-29. doi: 10.1016/j.steroids.2014.06.012. PubMed PMID: 24971815; PMCID: PMC4192010. 228. Hutcheon JA, Platt RW, Abrams B, Himes KP, Simhan HN, Bodnar LM. Pregnancy weight gain charts for obese and overweight women. Obesity (Silver Spring). 2015;23(3):532-5. doi: 10.1002/oby.21011. PubMed PMID: 25707378; PMCID: PMC4340088. 229. Hutcheon JA, Platt RW, Abrams B, Himes KP, Simhan HN, Bodnar LM. A weight-gain-for-gestational-age z score chart for the assessment of maternal weight gain in pregnancy. Am J Clin Nutr. 2013;97(5):1062-7. doi: 10.3945/ajcn.112.051706. PubMed PMID: 23466397; PMCID: PMC3625243. 230. Abeysena C, Jayawardana P. Sleep deprivation, physical activity and low income are risk factors for inadequate weight gain during pregnancy: a cohort study. J Obstet Gynaecol Res. 2011;37(7):734-40. doi: 10.1111/j.1447-0756.2010.01421.x. PubMed PMID: 21736667. 231. Hatcher KM, Smith RL, Chiang C, Li Z, Flaws JA, Mahoney MM. Association of phthalate exposure and endogenous hormones with self-reported sleep disruptions: results from the Midlife Women's Health Study. Menopause. 2020;27(11):1251-64. doi: 10.1097/GME.0000000000001614. PubMed PMID: 33110041. 232. Reeves KW, Santana MD, Manson JE, Hankinson SE, Zoeller RT, Bigelow C, Hou L, Wactawski-Wende J, Liu S, Tinker L, Calafat AM. Predictors of urinary phthalate biomarker concentrations in postmenopausal women. Environ Res. 2019;169:122-30. doi: 10.1016/j.envres.2018.10.024. PubMed PMID: 30447499; PMCID: PMC6347530. 233. VanderWeele TJ, Vansteelandt S. Mediation Analysis with Multiple Mediators. Epidemiol Methods. 2014;2(1):95-115. Epub 2015/01/13. doi: 10.1515/em-2012-0010. PubMed PMID: 25580377; PMCID: PMC4287269. 171 234. Naldi AC, Fayad PB, Prevost M, Sauve S. Analysis of steroid hormones and their conjugated forms in water and urine by on-line solid-phase extraction coupled to liquid chromatography tandem mass spectrometry. Chem Cent J. 2016;10:30. Epub 2016/05/10. doi: 10.1186/s13065-016-0174-z. PubMed PMID: 27158261; PMCID: PMC4859969. 235. Donangelo CMB, F.F. Pregnancy: Metabolic Adaptations and Nutritional Requirements. Caballero BF, P.M., Toldra, F., editor: Academic Press; 2016. 236. Lain KY, Catalano PM. Metabolic changes in pregnancy. Clin Obstet Gynecol. 2007;50(4):938-48. Epub 2007/11/06. doi: 10.1097/GRF.0b013e31815a5494. PubMed PMID: 17982337. 237. Pacyga DCC, A.M.; Braun, J.M.; Schantz, S.L.; Strakovsky, R.S. Phthalates/replacements are associated with maternal second-trimester glucose homeostasis and lipid-related metabolic factors. Environmental Health Perspectives. 2022(ISEE 2022 Abstracts: Strengthening the Global Role of Environmental Epidemiology). 238. Sampson M, Ling C, Sun Q, Harb R, Ashmaig M, Warnick R, Sethi A, Fleming JK, Otvos JD, Meeusen JW, Delaney SR, Jaffe AS, Shamburek R, Amar M, Remaley AT. A New Equation for Calculation of Low-Density Lipoprotein Cholesterol in Patients With Normolipidemia and/or Hypertriglyceridemia. JAMA Cardiol. 2020;5(5):540-8. Epub 2020/02/27. doi: 10.1001/jamacardio.2020.0013. PubMed PMID: 32101259; PMCID: PMC7240357 Abbott, Beckman, Siemens, Roche, ET Healthcare, Sphingotec, Brava, Quidel, Blade, and Novartis outside the submitted work. No other disclosures were reported. 239. Serrano SE, Braun J, Trasande L, Dills R, Sathyanarayana S. Phthalates and diet: a review of the food monitoring and epidemiology data. Environ Health. 2014;13(1):43. Epub 2014/06/05. doi: 10.1186/1476-069X-13-43. PubMed PMID: 24894065; PMCID: PMC4050989. 240. Pacyga DC, Sathyanarayana S, Strakovsky RS. Dietary Predictors of Phthalate and Bisphenol Exposures in Pregnant Women. Adv Nutr. 2019;10(5):803-15. doi: 10.1093/advances/nmz029. PubMed PMID: 31144713; PMCID: PMC6743849. 241. Chia AR, Chen LW, Lai JS, Wong CH, Neelakantan N, van Dam RM, Chong MF. Maternal Dietary Patterns and Birth Outcomes: A Systematic Review and Meta- Analysis. Adv Nutr. 2019;10(4):685-95. doi: 10.1093/advances/nmy123. PubMed PMID: 31041446; PMCID: PMC6628847. 242. Bragg M, Chavarro JE, Hamra GB, Hart JE, Tabb LP, Weisskopf MG, Volk HE, Lyall K. Prenatal Diet as a Modifier of Environmental Risk Factors for Autism and Related Neurodevelopmental Outcomes. Curr Environ Health Rep. 2022;9(2):324-38. doi: 10.1007/s40572-022-00347-7. PubMed PMID: 35305256; PMCID: PMC9098668. 172 243. Hennig B, Ettinger AS, Jandacek RJ, Koo S, McClain C, Seifried H, Silverstone A, Watkins B, Suk WA. Using nutrition for intervention and prevention against environmental chemical toxicity and associated diseases. Environ Health Perspect. 2007;115(4):493-5. doi: 10.1289/ehp.9549. PubMed PMID: 17450213; PMCID: PMC1852675. 244. Lecorguille M, Teo S, Phillips CM. Maternal Dietary Quality and Dietary Inflammation Associations with Offspring Growth, Placental Development, and DNA Methylation. Nutrients. 2021;13(9). doi: 10.3390/nu13093130. PubMed PMID: 34579008; PMCID: PMC8468062. 245. Song Q, Li R, Zhao Y, Zhu Q, Xia B, Chen S, Zhang Y. Evaluating effects of prenatal exposure to phthalates on neonatal birth weight: Structural equation model approaches. Chemosphere. 2018;205:674-81. Epub 2018/05/04. doi: 10.1016/j.chemosphere.2018.04.063. PubMed PMID: 29723725. 246. van den Dries MA, Keil AP, Tiemeier H, Pronk A, Spaan S, Santos S, Asimakopoulos AG, Kannan K, Gaillard R, Guxens M, Trasande L, Jaddoe VWV, Ferguson KK. Prenatal Exposure to Nonpersistent Chemical Mixtures and Fetal Growth: A Population-Based Study. Environ Health Perspect. 2021;129(11):117008. Epub 2021/11/25. doi: 10.1289/EHP9178. PubMed PMID: 34817287; PMCID: PMC8612241. 247. Pacyga DC, Talge NM, Gardiner JC, Calafat AM, Schantz SL, Strakovsky RS. Maternal diet quality moderates associations between parabens and birth outcomes. Environ Res. 2022;214(Pt 3):114078. Epub 2022/08/15. doi: 10.1016/j.envres.2022.114078. PubMed PMID: 35964672. 248. Pacyga DC, Henning M, Chiang C, Smith RL, Flaws JA, Strakovsky RS. Associations of Pregnancy History with BMI and Weight Gain in 45-54-Year-Old Women. Curr Dev Nutr. 2020;4(1):nzz139. doi: 10.1093/cdn/nzz139. PubMed PMID: 31893261; PMCID: PMC6933615. 249. Chiang C, Pacyga DC, Strakovsky RS, Smith RL, James-Todd T, Williams PL, Hauser R, Meling DD, Li Z, Flaws JA. Urinary phthalate metabolite concentrations and serum hormone levels in pre- and perimenopausal women from the Midlife Women's Health Study. Environ Int. 2021;156:106633. doi: 10.1016/j.envint.2021.106633. PubMed PMID: 34004451. 250. Warner GR, Pacyga DC, Strakovsky RS, Smith R, James-Todd T, Williams PL, Hauser R, Meling DD, Li Z, Flaws JA. Urinary phthalate metabolite concentrations and hot flashes in women from an urban convenience sample of midlife women. Environ Res. 2021;197:110891. doi: 10.1016/j.envres.2021.110891. PubMed PMID: 33722529; PMCID: PMC8187273. 251. Pacyga DC, Ryva BA, Nowak RA, Bulun SE, Yin P, Li Z, Flaws JA, Strakovsky RS. Midlife Urinary Phthalate Metabolite Concentrations and Prior Uterine Fibroid Diagnosis. Int J Environ Res Public Health. 2022;19(5). Epub 2022/03/11. doi: 10.3390/ijerph19052741. PubMed PMID: 35270433; PMCID: PMC8910544. 173 252. Lello S, Capozzi A, Scambia G. Osteoporosis and cardiovascular disease: an update. Gynecol Endocrinol. 2015;31(8):590-4. doi: 10.3109/09513590.2015.1041908. PubMed PMID: 26036806. 253. Perng W, Kasper NM, Watkins DJ, Sanchez BN, Meeker JD, Cantoral A, Solano- González M, Tellez-Rojo MM, Peterson K. Exposure to Endocrine-Disrupting Chemicals During Pregnancy Is Associated with Weight Change Through 1 Year Postpartum Among Women in the Early-Life Exposure in Mexico to Environmental Toxicants Project. J Womens Health (Larchmt). 2020. Epub 2020/04/03. doi: 10.1089/jwh.2019.8078. PubMed PMID: 32233978. 254. Cao XL, Zhao W, Churchill R, Hilts C. Occurrence of Di-(2-ethylhexyl) adipate and phthalate plasticizers in samples of meat, fish, and cheese and their packaging films. J Food Prot. 2014;77(4):610-20. Epub 2014/04/01. doi: 10.4315/0362-028X.JFP- 13-380. PubMed PMID: 24680073. 255. (CDC) CfDCaP. Fourth National Report on Human Exposure to Environmental Chemicals, Updated Tables (March 2018). Atlanta, Georgia: U.S. Department of Health and Human Services, Centers for Disease Control and Prevention, 2018. 256. CDC CfDCaPC. Biomonitoring Summary 2016 [updated 01/13/2019]. Available from: https://www.cdc.gov/biomonitoring/biomonitoring_summaries.html. 257. Erythropel HC, Maric M, Nicell JA, Leask RL, Yargeau V. Leaching of the plasticizer di(2-ethylhexyl)phthalate (DEHP) from plastic containers and the question of human exposure. Appl Microbiol Biotechnol. 2014;98(24):9967-81. Epub 2014/11/08. doi: 10.1007/s00253-014-6183-8. PubMed PMID: 25376446. 258. Philippat C, Wolff MS, Calafat AM, Ye X, Bausell R, Meadows M, Stone J, Slama R, Engel SM. Prenatal exposure to environmental phenols: concentrations in amniotic fluid and variability in urinary concentrations during pregnancy. Environ Health Perspect. 2013;121(10):1225-31. doi: 10.1289/ehp.1206335. PubMed PMID: 23942273; PMCID: PMC3801458. 259. Rochester JR, Bolden AL. Bisphenol S and F: A Systematic Review and Comparison of the Hormonal Activity of Bisphenol A Substitutes. Environ Health Perspect. 2015;123(7):643-50. Epub 2015/03/17. doi: 10.1289/ehp.1408989. PubMed PMID: 25775505; PMCID: PMC4492270. 260. Lorber M, Schecter A, Paepke O, Shropshire W, Christensen K, Birnbaum L. Exposure assessment of adult intake of bisphenol A (BPA) with emphasis on canned food dietary exposures. Environ Int. 2015;77:55-62. Epub 2015/02/04. doi: 10.1016/j.envint.2015.01.008. PubMed PMID: 25645382; PMCID: PMC4469126. 261. Noonan GO, Ackerman LK, Begley TH. Concentration of bisphenol A in highly consumed canned foods on the U.S. market. J Agric Food Chem. 2011;59(13):7178-85. Epub 2011/05/24. doi: 10.1021/jf201076f. PubMed PMID: 21598963. 174 262. Calafat AM, Longnecker MP, Koch HM, Swan SH, Hauser R, Goldman LR, Lanphear BP, Rudel RA, Engel SM, Teitelbaum SL, Whyatt RM, Wolff MS. Optimal Exposure Biomarkers for Nonpersistent Chemicals in Environmental Epidemiology. Environ Health Perspect. 2015;123(7):A166-8. Epub 2015/07/02. doi: 10.1289/ehp.1510041. PubMed PMID: 26132373; PMCID: PMC4492274. 263. Chandra A, Martinez GM, Mosher WD, Abma JC, Jones J. Fertility, family planning, and reproductive health of U.S. women: data from the 2002 National Survey of Family Growth. Vital Health Stat 23. 2005(25):1-160. PubMed PMID: 16532609. 264. (CDC) CoDCaP. National Health and Nutrition Examination Survey Data. In: (NCHS) NCfHS, editor. Hyattsville, Maryland: U.S. Department of Health and Human Services, Centers for Disease Control and Prevention; 2013-2014. 265. Aker AM, Watkins DJ, Johns LE, Ferguson KK, Soldin OP, Anzalota Del Toro LV, Alshawabkeh AN, Cordero JF, Meeker JD. Phenols and parabens in relation to reproductive and thyroid hormones in pregnant women. Environ Res. 2016;151:30-7. Epub 2016/10/21. doi: 10.1016/j.envres.2016.07.002. PubMed PMID: 27448730; PMCID: PMC5071140. 266. Huang PC, Tsai CH, Liang WY, Li SS, Huang HB, Kuo PL. Early Phthalates Exposure in Pregnant Women Is Associated with Alteration of Thyroid Hormones. PLoS One. 2016;11(7):e0159398. doi: 10.1371/journal.pone.0159398. PubMed PMID: 27455052; PMCID: PMC4959782. 267. Karwacka A, Zamkowska D, Radwan M, Jurewicz J. Exposure to modern, widespread environmental endocrine disrupting chemicals and their effect on the reproductive potential of women: an overview of current epidemiological evidence. Hum Fertil (Camb). 2017:1-24. Epub 2017/08/02. doi: 10.1080/14647273.2017.1358828. PubMed PMID: 28758506. 268. Findlay LC, Kohen DE. Bisphenol A and child and youth behaviour: Canadian Health Measures Survey 2007 to 2011. Health Rep. 2015;26(8):3-9. Epub 2015/08/20. PubMed PMID: 26288316. 269. Buckley JP, Quiros-Alcala L, Teitelbaum SL, Calafat AM, Wolff MS, Engel SM. Associations of prenatal environmental phenol and phthalate biomarkers with respiratory and allergic diseases among children aged 6 and 7years. Environ Int. 2018;115:79-88. Epub 2018/03/20. doi: 10.1016/j.envint.2018.03.016. PubMed PMID: 29550712; PMCID: PMC5970077. 270. Bhandari R, Xiao J, Shankar A. Urinary bisphenol A and obesity in U.S. children. Am J Epidemiol. 2013;177(11):1263-70. Epub 2013/04/06. doi: 10.1093/aje/kws391. PubMed PMID: 23558351; PMCID: PMC3664337. 271. Deierlein AL, Wolff MS, Pajak A, Pinney SM, Windham GC, Galvez MP, Rybak M, Calafat AM, Kushi LH, Biro FM, Teitelbaum SL. Phenol Concentrations During Childhood and Subsequent Measures of Adiposity Among Young Girls. Am J Epidemiol. 175 2017;186(5):581-92. Epub 2017/05/20. doi: 10.1093/aje/kwx136. PubMed PMID: 28525533; PMCID: PMC5600702. 272. Lakind JS, Naiman DQ. Daily intake of bisphenol A and potential sources of exposure: 2005-2006 National Health and Nutrition Examination Survey. J Expo Sci Environ Epidemiol. 2011;21(3):272-9. Epub 2010/03/20. doi: 10.1038/jes.2010.9. PubMed PMID: 20237498; PMCID: PMC3079892. 273. Fromme H, Gruber L, Schlummer M, Wolz G, Bohmer S, Angerer J, Mayer R, Liebl B, Bolte G. Intake of phthalates and di(2-ethylhexyl)adipate: results of the Integrated Exposure Assessment Survey based on duplicate diet samples and biomonitoring data. Environ Int. 2007;33(8):1012-20. Epub 2007/07/06. doi: 10.1016/j.envint.2007.05.006. PubMed PMID: 17610953. 274. Muncke J. Endocrine disrupting chemicals and other substances of concern in food contact materials: an updated review of exposure, effect and risk assessment. J Steroid Biochem Mol Biol. 2011;127(1-2):118-27. Epub 2010/11/16. doi: 10.1016/j.jsbmb.2010.10.004. PubMed PMID: 21073950. 275. Muncke J, Backhaus T, Geueke B, Maffini MV, Martin OV, Myers JP, Soto AM, Trasande L, Trier X, Scheringer M. Scientific Challenges in the Risk Assessment of Food Contact Materials. Environ Health Perspect. 2017;125(9):095001. Epub 2017/09/13. doi: 10.1289/EHP644. PubMed PMID: 28893723; PMCID: PMC5915200. 276. Wittassek M, Koch HM, Angerer J, Bruning T. Assessing exposure to phthalates - the human biomonitoring approach. Mol Nutr Food Res. 2011;55(1):7-31. Epub 2010/06/22. doi: 10.1002/mnfr.201000121. PubMed PMID: 20564479. 277. Abduljalil K, Furness P, Johnson TN, Rostami-Hodjegan A, Soltani H. Anatomical, physiological and metabolic changes with gestational age during normal pregnancy: a database for parameters required in physiologically based pharmacokinetic modelling. Clin Pharmacokinet. 2012;51(6):365-96. Epub 2012/04/21. doi: 10.2165/11597440-000000000-00000. PubMed PMID: 22515555. 278. Tasnif Y, Morado J, Hebert MF. Pregnancy-related pharmacokinetic changes. Clin Pharmacol Ther. 2016;100(1):53-62. Epub 2016/04/16. doi: 10.1002/cpt.382. PubMed PMID: 27082931. 279. Chen LW, Low YL, Fok D, Han WM, Chong YS, Gluckman P, Godfrey K, Kwek K, Saw SM, Soh SE, Tan KH, Chong MF, van Dam RM. Dietary changes during pregnancy and the postpartum period in Singaporean Chinese, Malay and Indian women: the GUSTO birth cohort study. Public Health Nutr. 2014;17(9):1930-8. Epub 2013/06/29. doi: 10.1017/S1368980013001730. PubMed PMID: 23806144. 280. Rifas-Shiman SL, Rich-Edwards JW, Willett WC, Kleinman KP, Oken E, Gillman MW. Changes in dietary intake from the first to the second trimester of pregnancy. Paediatr Perinat Epidemiol. 2006;20(1):35-42. Epub 2006/01/20. doi: 10.1111/j.1365- 3016.2006.00691.x. PubMed PMID: 16420339; PMCID: PMC1488723. 176 281. Casas M, Valvi D, Luque N, Ballesteros-Gomez A, Carsin AE, Fernandez MF, Koch HM, Mendez MA, Sunyer J, Rubio S, Vrijheid M. Dietary and sociodemographic determinants of bisphenol A urine concentrations in pregnant women and children. Environ Int. 2013;56:10-8. Epub 2013/04/02. doi: 10.1016/j.envint.2013.02.014. PubMed PMID: 23542682. 282. Serrano SE, Karr CJ, Seixas NS, Nguyen RH, Barrett ES, Janssen S, Redmon B, Swan SH, Sathyanarayana S. Dietary phthalate exposure in pregnant women and the impact of consumer practices. Int J Environ Res Public Health. 2014;11(6):6193- 215. Epub 2014/06/14. doi: 10.3390/ijerph110606193. PubMed PMID: 24927036; PMCID: PMC4078574. 283. Quiros-Alcala L, Eskenazi B, Bradman A, Ye X, Calafat AM, Harley K. Determinants of urinary bisphenol A concentrations in Mexican/Mexican--American pregnant women. Environ Int. 2013;59:152-60. Epub 2013/07/03. doi: 10.1016/j.envint.2013.05.016. PubMed PMID: 23816546; PMCID: PMC3954740. 284. Braun JM, Kalkbrenner AE, Calafat AM, Bernert JT, Ye X, Silva MJ, Barr DB, Sathyanarayana S, Lanphear BP. Variability and predictors of urinary bisphenol A concentrations during pregnancy. Environ Health Perspect. 2011;119(1):131-7. Epub 2011/01/06. doi: 10.1289/ehp.1002366. PubMed PMID: 21205581; PMCID: PMC3018492. 285. Callan AC, Hinwood AL, Heffernan A, Eaglesham G, Mueller J, Odland JO. Urinary bisphenol A concentrations in pregnant women. Int J Hyg Environ Health. 2013;216(6):641-4. Epub 2012/11/15. doi: 10.1016/j.ijheh.2012.10.002. PubMed PMID: 23149244. 286. Calafat AM, Ye X, Wong LY, Bishop AM, Needham LL. Urinary concentrations of four parabens in the U.S. population: NHANES 2005-2006. Environ Health Perspect. 2010;118(5):679-85. Epub 2010/01/09. doi: 10.1289/ehp.0901560. PubMed PMID: 20056562; PMCID: PMC2866685. 287. Calafat AM, Ye X, Wong LY, Reidy JA, Needham LL. Urinary concentrations of triclosan in the U.S. population: 2003-2004. Environ Health Perspect. 2008;116(3):303- 7. Epub 2008/03/13. doi: 10.1289/ehp.10768. PubMed PMID: 18335095; PMCID: PMC2265044. 288. (FDA) USFaDA. Science and Research (Food) 2018 [01/13/2019]. Available from: https://www.fda.gov/Food/FoodScienceResearch/default.htm. 289. (EFSA) EFSA. Chemical Contaminants [01/13/2019]. Available from: http://www.efsa.europa.eu/en/topics/topic/chemical-contaminants. 290. Careghini A, Mastorgio AF, Saponaro S, Sezenna E. Bisphenol A, nonylphenols, benzophenones, and benzotriazoles in soils, groundwater, surface water, sediments, and food: a review. Environ Sci Pollut Res Int. 2015;22(8):5711-41. Epub 2014/12/31. doi: 10.1007/s11356-014-3974-5. PubMed PMID: 25548011; PMCID: PMC4381092. 177 291. Rudel RA, Gray JM, Engel CL, Rawsthorne TW, Dodson RE, Ackerman JM, Rizzo J, Nudelman JL, Brody JG. Food packaging and bisphenol A and bis(2-ethyhexyl) phthalate exposure: findings from a dietary intervention. Environ Health Perspect. 2011;119(7):914-20. Epub 2011/04/01. doi: 10.1289/ehp.1003170. PubMed PMID: 21450549; PMCID: PMC3223004. 292. Cao XL. Phthalate Esters in Foods: Sources, Occurrence, and Analytical Methods. Comprehensive Reviews in Food Science and Food Safety. 2010;9(1):21-43. 293. Sakhi AK, Lillegaard IT, Voorspoels S, Carlsen MH, Loken EB, Brantsaeter AL, Haugen M, Meltzer HM, Thomsen C. Concentrations of phthalates and bisphenol A in Norwegian foods and beverages and estimated dietary exposure in adults. Environ Int. 2014;73:259-69. Epub 2014/09/01. doi: 10.1016/j.envint.2014.08.005. PubMed PMID: 25173060. 294. Van Holderbeke M, Geerts L, Vanermen G, Servaes K, Sioen I, De Henauw S, Fierens T. Determination of contamination pathways of phthalates in food products sold on the Belgian market. Environ Res. 2014;134:345-52. Epub 2014/09/10. doi: 10.1016/j.envres.2014.08.012. PubMed PMID: 25203818. 295. Schecter A, Lorber M, Guo Y, Wu Q, Yun SH, Kannan K, Hommel M, Imran N, Hynan LS, Cheng D, Colacino JA, Birnbaum LS. Phthalate concentrations and dietary exposure from food purchased in New York State. Environ Health Perspect. 2013;121(4):473-94. Epub 2013/03/07. doi: 10.1289/ehp.1206367. PubMed PMID: 23461894; PMCID: PMC3620091. 296. Guo Y, Zhang Z, Liu L, Li Y, Ren N, Kannan K. Occurrence and profiles of phthalates in foodstuffs from China and their implications for human exposure. J Agric Food Chem. 2012;60(27):6913-9. Epub 2012/06/19. doi: 10.1021/jf3021128. PubMed PMID: 22703192. 297. Liao C, Kannan K. Concentrations and profiles of bisphenol A and other bisphenol analogues in foodstuffs from the United States and their implications for human exposure. J Agric Food Chem. 2013;61(19):4655-62. Epub 2013/04/26. doi: 10.1021/jf400445n. PubMed PMID: 23614805. 298. Cao XL, Perez-Locas C, Dufresne G, Clement G, Popovic S, Beraldin F, Dabeka RW, Feeley M. Concentrations of bisphenol A in the composite food samples from the 2008 Canadian total diet study in Quebec City and dietary intake estimates. Food Addit Contam Part A Chem Anal Control Expo Risk Assess. 2011;28(6):791-8. Epub 2011/05/31. doi: 10.1080/19440049.2010.513015. PubMed PMID: 21623504; PMCID: PMC3118530. 299. Bradley EL, Burden RA, Bentayeb K, Driffield M, Harmer N, Mortimer DN, Speck DR, Ticha J, Castle L. Exposure to phthalic acid, phthalate diesters and phthalate monoesters from foodstuffs: UK total diet study results. Food Addit Contam Part A Chem Anal Control Expo Risk Assess. 2013;30(4):735-42. Epub 2013/05/07. doi: 10.1080/19440049.2013.781684. PubMed PMID: 23641808. 178 300. Schecter A, Malik N, Haffner D, Smith S, Harris TR, Paepke O, Birnbaum L. Bisphenol A (BPA) in U.S. food. Environ Sci Technol. 2010;44(24):9425-30. Epub 2010/11/03. doi: 10.1021/es102785d. PubMed PMID: 21038926. 301. Mariscal-Arcas M, Rivas A, Granada A, Monteagudo C, Murcia MA, Olea- Serrano F. Dietary exposure assessment of pregnant women to bisphenol-A from cans and microwave containers in Southern Spain. Food Chem Toxicol. 2009;47(2):506-10. Epub 2009/01/06. doi: 10.1016/j.fct.2008.12.011. PubMed PMID: 19121362. 302. Bai PY, Wittert GA, Taylor AW, Martin SA, Milne RW, Shi Z. The association of socio-demographic status, lifestyle factors and dietary patterns with total urinary phthalates in Australian men. PLoS One. 2015;10(4):e0122140. Epub 2015/04/16. doi: 10.1371/journal.pone.0122140. PubMed PMID: 25875472; PMCID: PMC4398403. 303. Mervish N, McGovern KJ, Teitelbaum SL, Pinney SM, Windham GC, Biro FM, Kushi LH, Silva MJ, Ye X, Calafat AM, Wolff MS, Bcerp. Dietary predictors of urinary environmental biomarkers in young girls, BCERP, 2004-7. Environ Res. 2014;133:12-9. Epub 2014/06/07. doi: 10.1016/j.envres.2014.04.040. PubMed PMID: 24906063; PMCID: PMC4119560. 304. Martine B, Marie-Jeanne T, Cendrine D, Fabrice A, Marc C. Assessment of adult human exposure to phthalate esters in the urban centre of Paris (France). Bull Environ Contam Toxicol. 2013;90(1):91-6. Epub 2012/10/24. doi: 10.1007/s00128-012-0859-5. PubMed PMID: 23090363. 305. Cacho JI, Campillo N, Vinas P, Hernandez-Cordoba M. Determination of alkylphenols and phthalate esters in vegetables and migration studies from their packages by means of stir bar sorptive extraction coupled to gas chromatography-mass spectrometry. J Chromatogr A. 2012;1241:21-7. Epub 2012/04/27. doi: 10.1016/j.chroma.2012.04.018. PubMed PMID: 22533911. 306. Cao XL, Corriveau J, Popovic S. Bisphenol a in canned food products from canadian markets. J Food Prot. 2010;73(6):1085-9. Epub 2010/06/12. PubMed PMID: 20537264. 307. Lu J, Wu J, Stoffella PJ, Wilson PC. Analysis of bisphenol A, nonylphenol, and natural estrogens in vegetables and fruits using gas chromatography-tandem mass spectrometry. J Agric Food Chem. 2013;61(1):84-9. Epub 2012/12/12. doi: 10.1021/jf304971k. PubMed PMID: 23215552. 308. Fierens T, Van Holderbeke M, Willems H, De Henauw S, Sioen I. Transfer of eight phthalates through the milk chain--a case study. Environ Int. 2013;51:1-7. Epub 2012/11/10. doi: 10.1016/j.envint.2012.10.002. PubMed PMID: 23138015. 309. Beltifa A, Feriani A, Machreki M, Ghorbel A, Ghazouani L, Di Bella G, Van Loco J, Reyns T, Mansour HB. Plasticizers and bisphenol A, in packaged foods sold in the Tunisian markets: study of their acute in vivo toxicity and their environmental fate. 179 Environ Sci Pollut Res Int. 2017;24(28):22382-92. Epub 2017/08/13. doi: 10.1007/s11356-017-9861-0. PubMed PMID: 28801775. 310. (EFSA) EFSA. Scientific Opinion on the risks to public health related to the presence of bisphenol A (BPA) in foodstuffs – Part: exposure assessment. EFSA Panel on Food Contact Materials, Enzymes, Flavourings and Processing Aids (CEF): 2013. 311. Shao B, Han H, Tu X, Huang L. Analysis of alkylphenol and bisphenol A in eggs and milk by matrix solid phase dispersion extraction and liquid chromatography with tandem mass spectrometry. J Chromatogr B Analyt Technol Biomed Life Sci. 2007;850(1-2):412-6. Epub 2007/02/03. doi: 10.1016/j.jchromb.2006.12.033. PubMed PMID: 17270504. 312. Sajiki J, Miyamoto F, Fukata H, Mori C, Yonekubo J, Hayakawa K. Bisphenol A (BPA) and its source in foods in Japanese markets. Food Addit Contam. 2007;24(1):103-12. Epub 2006/12/14. doi: 10.1080/02652030600936383. PubMed PMID: 17164221. 313. Bemrah N, Jean J, Riviere G, Sanaa M, Leconte S, Bachelot M, Deceuninck Y, Bizec BL, Dauchy X, Roudot AC, Camel V, Grob K, Feidt C, Picard-Hagen N, Badot PM, Foures F, Leblanc JC. Assessment of dietary exposure to bisphenol A in the French population with a special focus on risk characterisation for pregnant French women. Food Chem Toxicol. 2014;72:90-7. Epub 2014/07/12. doi: 10.1016/j.fct.2014.07.005. PubMed PMID: 25014159. 314. Rosenmai AK, Bengtstrom L, Taxvig C, Trier X, Petersen JH, Svingen T, Binderup ML, Barbara Medea Alice VV, Dybdahl M, Granby K, Vinggaard AM. An effect-directed strategy for characterizing emerging chemicals in food contact materials made from paper and board. Food Chem Toxicol. 2017;106(Pt A):250-9. Epub 2017/06/03. doi: 10.1016/j.fct.2017.05.061. PubMed PMID: 28571769. 315. Varshavsky JR, Morello-Frosch R, Woodruff TJ, Zota AR. Dietary sources of cumulative phthalates exposure among the U.S. general population in NHANES 2005- 2014. Environ Int. 2018;115:417-29. Epub 2018/04/02. doi: 10.1016/j.envint.2018.02.029. PubMed PMID: 29605141; PMCID: PMC5970069. 316. Zota AR, Phillips CA, Mitro SD. Recent Fast Food Consumption and Bisphenol A and Phthalates Exposures among the U.S. Population in NHANES, 2003-2010. Environ Health Perspect. 2016;124(10):1521-8. Epub 2016/04/14. doi: 10.1289/ehp.1510803. PubMed PMID: 27072648; PMCID: PMC5047792 interests and that their freedom to design, conduct, interpret, and publish research is not compromised by any controlling sponsor. 317. Tordjman K, Grinshpan L, Novack L, Goen T, Segev D, Beacher L, Stern N, Berman T. Exposure to endocrine disrupting chemicals among residents of a rural vegetarian/vegan community. Environ Int. 2016;97:68-75. Epub 2016/10/30. doi: 10.1016/j.envint.2016.10.018. PubMed PMID: 27792907. 180 318. Sathyanarayana S, Alcedo G, Saelens BE, Zhou C, Dills RL, Yu J, Lanphear B. Unexpected results in a randomized dietary trial to reduce phthalate and bisphenol A exposures. J Expo Sci Environ Epidemiol. 2013;23(4):378-84. Epub 2013/02/28. doi: 10.1038/jes.2013.9. PubMed PMID: 23443238. 319. Moreira MA, Andre LC, Cardeal ZL. Analysis of phthalate migration to food simulants in plastic containers during microwave operations. Int J Environ Res Public Health. 2013;11(1):507-26. Epub 2014/01/02. doi: 10.3390/ijerph110100507. PubMed PMID: 24380980; PMCID: PMC3924457. 320. Li C, Xu J, Chen D, Xiao Y. Detection of phthalates migration from disposable tablewares to drinking water using hexafluoroisopropanol-induced catanionic surfactant coacervate extraction. J Pharm Anal. 2016;6(5):292-9. Epub 2016/10/01. doi: 10.1016/j.jpha.2016.04.002. PubMed PMID: 29403995; PMCID: PMC5762623. 321. Cooper JE, Kendig EL, Belcher SM. Assessment of bisphenol A released from reusable plastic, aluminium and stainless steel water bottles. Chemosphere. 2011;85(6):943-7. Epub 2011/07/12. doi: 10.1016/j.chemosphere.2011.06.060. PubMed PMID: 21741673; PMCID: PMC3210908. 322. Carwile JL, Luu HT, Bassett LS, Driscoll DA, Yuan C, Chang JY, Ye X, Calafat AM, Michels KB. Polycarbonate bottle use and urinary bisphenol A concentrations. Environ Health Perspect. 2009;117(9):1368-72. Epub 2009/09/15. doi: 10.1289/ehp.0900604. PubMed PMID: 19750099; PMCID: PMC2737011. 323. Makris KC, Andra SS, Jia A, Herrick L, Christophi CA, Snyder SA, Hauser R. Association between water consumption from polycarbonate containers and bisphenol A intake during harsh environmental conditions in summer. Environ Sci Technol. 2013;47(7):3333-43. Epub 2013/03/02. doi: 10.1021/es304038k. PubMed PMID: 23448553. 324. Braunrath RP, D.; Padlesak, S.; Cichna-Markl, M. Determination of Bisphenol A in Canned Foods by Immunoaffinity Chromatography, HPLC, and Fluorescence Detection. Journal of Agricultural and Food Chemistry. 2005;53(23):8911-7. doi: DOI: 10.1021/jf051525j 325. Geens T, Apelbaum TZ, Goeyens L, Neels H, Covaci A. Intake of bisphenol A from canned beverages and foods on the Belgian market. Food Addit Contam Part A Chem Anal Control Expo Risk Assess. 2010;27(11):1627-37. Epub 2010/09/14. doi: 10.1080/19440049.2010.508183. PubMed PMID: 20835936. 326. Tzatzarakis MN, Karzi V, Vakonaki E, Goumenou M, Kavvalakis M, Stivaktakis P, Tsitsimpikou C, Tsakiris I, Rizos AK, Tsatsakis AM. Bisphenol A in soft drinks and canned foods and data evaluation. Food Addit Contam Part B Surveill. 2017;10(2):85- 90. Epub 2016/11/30. doi: 10.1080/19393210.2016.1266522. PubMed PMID: 27897085. 181 327. CDC CfDCaPC. Fourth National Report on Human Exposure to Environmental Chemicals. Atlanta, Georgia: 2018. 328. Adibi JJ, Whyatt RM, Williams PL, Calafat AM, Camann D, Herrick R, Nelson H, Bhat HK, Perera FP, Silva MJ, Hauser R. Characterization of phthalate exposure among pregnant women assessed by repeat air and urine samples. Environ Health Perspect. 2008;116(4):467-73. Epub 2008/04/17. doi: 10.1289/ehp.10749. PubMed PMID: 18414628; PMCID: 2291011. 329. Fisher M, Arbuckle TE, Mallick R, LeBlanc A, Hauser R, Feeley M, Koniecki D, Ramsay T, Provencher G, Berube R, Walker M. Bisphenol A and phthalate metabolite urinary concentrations: Daily and across pregnancy variability. J Expo Sci Environ Epidemiol. 2015;25(3):231-9. Epub 2014/09/25. doi: 10.1038/jes.2014.65. PubMed PMID: 25248937; PMCID: PMC4408490. 330. Smith KW, Braun JM, Williams PL, Ehrlich S, Correia KF, Calafat AM, Ye X, Ford J, Keller M, Meeker JD, Hauser R. Predictors and variability of urinary paraben concentrations in men and women, including before and during pregnancy. Environ Health Perspect. 2012;120(11):1538-43. Epub 2012/06/23. doi: 10.1289/ehp.1104614. PubMed PMID: 22721761; PMCID: PMC3556607. 331. Meeker JD, Cantonwine DE, Rivera-Gonzalez LO, Ferguson KK, Mukherjee B, Calafat AM, Ye X, Anzalota Del Toro LV, Crespo-Hernandez N, Jimenez-Velez B, Alshawabkeh AN, Cordero JF. Distribution, variability, and predictors of urinary concentrations of phenols and parabens among pregnant women in Puerto Rico. Environ Sci Technol. 2013;47(7):3439-47. Epub 2013/03/09. doi: 10.1021/es400510g. PubMed PMID: 23469879; PMCID: PMC3638245. 332. McGowan CA, Curran S, McAuliffe FM. Relative validity of a food frequency questionnaire to assess nutrient intake in pregnant women. J Hum Nutr Diet. 2014;27 Suppl 2:167-74. Epub 2013/05/01. doi: 10.1111/jhn.12120. PubMed PMID: 23627971. 333. Venter C, Higgins B, Grundy J, Clayton CB, Gant C, Dean T. Reliability and validity of a maternal food frequency questionnaire designed to estimate consumption of common food allergens. J Hum Nutr Diet. 2006;19(2):129-38. Epub 2006/03/15. doi: 10.1111/j.1365-277X.2006.00677.x. PubMed PMID: 16533375. 334. Manuck TA, Rice MM, Bailit JL, Grobman WA, Reddy UM, Wapner RJ, Thorp JM, Caritis SN, Prasad M, Tita AT, Saade GR, Sorokin Y, Rouse DJ, Blackwell SC, Tolosa JE, Eunice Kennedy Shriver National Institute of Child H, Human Development Maternal-Fetal Medicine Units N. Preterm neonatal morbidity and mortality by gestational age: a contemporary cohort. Am J Obstet Gynecol. 2016;215(1):103 e1- e14. doi: 10.1016/j.ajog.2016.01.004. PubMed PMID: 26772790; PMCID: PMC4921282. 335. Markopoulou P, Papanikolaou E, Analytis A, Zoumakis E, Siahanidou T. Preterm Birth as a Risk Factor for Metabolic Syndrome and Cardiovascular Disease in Adult Life: 182 A Systematic Review and Meta-Analysis. J Pediatr. 2019;210:69-80 e5. doi: 10.1016/j.jpeds.2019.02.041. PubMed PMID: 30992219. 336. Spong CY. Defining "term" pregnancy: recommendations from the Defining "Term" Pregnancy Workgroup. JAMA. 2013;309(23):2445-6. Epub 2013/05/07. doi: 10.1001/jama.2013.6235. PubMed PMID: 23645117. 337. Espel EV, Glynn LM, Sandman CA, Davis EP. Longer gestation among children born full term influences cognitive and motor development. PLoS One. 2014;9(11):e113758. Epub 2014/11/26. doi: 10.1371/journal.pone.0113758. PubMed PMID: 25423150; PMCID: PMC4244187. 338. Tita AT, Landon MB, Spong CY, Lai Y, Leveno KJ, Varner MW, Moawad AH, Caritis SN, Meis PJ, Wapner RJ, Sorokin Y, Miodovnik M, Carpenter M, Peaceman AM, O'Sullivan MJ, Sibai BM, Langer O, Thorp JM, Ramin SM, Mercer BM, Eunice Kennedy Shriver NM-FMUN. Timing of elective repeat cesarean delivery at term and neonatal outcomes. N Engl J Med. 2009;360(2):111-20. Epub 2009/01/09. doi: 10.1056/NEJMoa0803267. PubMed PMID: 19129525; PMCID: PMC2811696. 339. Mitchell EA, Stewart AW, Braithwaite I, Hancox RJ, Murphy R, Wall C, Beasley R, Group IPTS. Birth weight and subsequent body mass index in children: an international cross-sectional study. Pediatr Obes. 2017;12(4):280-5. doi: 10.1111/ijpo.12138. PubMed PMID: 27170099. 340. Zhao Y, Wang SF, Mu M, Sheng J. Birth weight and overweight/obesity in adults: a meta-analysis. Eur J Pediatr. 2012;171(12):1737-46. doi: 10.1007/s00431-012-1701- 0. PubMed PMID: 22383072. 341. Bjorstad AR, Irgens-Hansen K, Daltveit AK, Irgens LM. Macrosomia: mode of delivery and pregnancy outcome. Acta Obstet Gynecol Scand. 2010;89(5):664-9. Epub 2010/03/20. doi: 10.3109/00016341003686099. PubMed PMID: 20235897. 342. Lindley AA, Benson JE, Grimes C, Cole TM, 3rd, Herman AA. The relationship in neonates between clinically measured head circumference and brain volume estimated from head CT-scans. Early Hum Dev. 1999;56(1):17-29. Epub 1999/10/26. doi: 10.1016/s0378-3782(99)00033-x. PubMed PMID: 10530903. 343. Mortensen ME, Calafat AM, Ye X, Wong LY, Wright DJ, Pirkle JL, Merrill LS, Moye J. Urinary concentrations of environmental phenols in pregnant women in a pilot study of the National Children's Study. Environ Res. 2014;129:32-8. Epub 2014/02/18. doi: 10.1016/j.envres.2013.12.004. PubMed PMID: 24529000; PMCID: PMC4530794. 344. Program NB. Parabens 2017 [cited 2019 09/04]. Available from: https://www.cdc.gov/biomonitoring/Parabens_BiomonitoringSummary.html. 345. Bereketoglu C, Pradhan A. Comparative transcriptional analysis of methylparaben and propylparaben in zebrafish. Sci Total Environ. 2019;671:129-39. doi: 10.1016/j.scitotenv.2019.03.358. PubMed PMID: 30928742. 183 346. Watkins DJ, Ferguson KK, Anzalota Del Toro LV, Alshawabkeh AN, Cordero JF, Meeker JD. Associations between urinary phenol and paraben concentrations and markers of oxidative stress and inflammation among pregnant women in Puerto Rico. Int J Hyg Environ Health. 2015;218(2):212-9. Epub 2014/12/02. doi: 10.1016/j.ijheh.2014.11.001. PubMed PMID: 25435060; PMCID: PMC4323928. 347. Aung MT, Ferguson KK, Cantonwine DE, Bakulski KM, Mukherjee B, Loch- Caruso R, McElrath TF, Meeker JD. Associations between maternal plasma measurements of inflammatory markers and urinary levels of phenols and parabens during pregnancy: A repeated measures study. Sci Total Environ. 2019;650(Pt 1):1131- 40. Epub 2018/10/13. doi: 10.1016/j.scitotenv.2018.08.356. PubMed PMID: 30308801; PMCID: PMC6236678. 348. Bairati C, Goi G, Lombardo A, Tettamanti G. The esters of p-hydroxy-benzoate (parabens) inhibit the release of lysosomal enzymes by mitogen-stimulated peripheral human lymphocytes in culture. Clin Chim Acta. 1994;224(2):147-57. doi: 10.1016/0009- 8981(94)90181-3. PubMed PMID: 8004785. 349. Guzel Bayulken D, Ayaz Tuylu B. In vitro genotoxic and cytotoxic effects of some paraben esters on human peripheral lymphocytes. Drug Chem Toxicol. 2019;42(4):386- 93. doi: 10.1080/01480545.2018.1457049. PubMed PMID: 29681198. 350. Aker AM, Johns L, McElrath TF, Cantonwine DE, Mukherjee B, Meeker JD. Associations between maternal phenol and paraben urinary biomarkers and maternal hormones during pregnancy: A repeated measures study. Environ Int. 2018. Epub 2018/01/26. doi: 10.1016/j.envint.2018.01.006. PubMed PMID: 29366524. 351. Kolatorova L, Vitku J, Hampl R, Adamcova K, Skodova T, Simkova M, Parizek A, Starka L, Duskova M. Exposure to bisphenols and parabens during pregnancy and relations to steroid changes. Environ Res. 2018;163:115-22. doi: 10.1016/j.envres.2018.01.031. PubMed PMID: 29433019. 352. Ahn HJ, An BS, Jung EM, Yang H, Choi KC, Jeung EB. Parabens inhibit the early phase of folliculogenesis and steroidogenesis in the ovaries of neonatal rats. Mol Reprod Dev. 2012;79(9):626-36. Epub 2012/07/11. doi: 10.1002/mrd.22070. PubMed PMID: 22777679. 353. Lee JH, Lee M, Ahn C, Kang HY, Tran DN, Jeung EB. Parabens Accelerate Ovarian Dysfunction in a 4-Vinylcyclohexene Diepoxide-Induced Ovarian Failure Model. Int J Environ Res Public Health. 2017;14(2). Epub 2017/02/18. doi: 10.3390/ijerph14020161. PubMed PMID: 28208728; PMCID: PMC5334715. 354. Elmore SE, Cano-Sancho G, La Merrill MA. Disruption of normal adipocyte development and function by methyl- and propyl- paraben exposure. Toxicol Lett. 2020;334:27-35. doi: 10.1016/j.toxlet.2020.09.009. PubMed PMID: 32956827. 184 355. Hu P, Chen X, Whitener RJ, Boder ET, Jones JO, Porollo A, Chen J, Zhao L. Effects of parabens on adipocyte differentiation. Toxicol Sci. 2013;131(1):56-70. doi: 10.1093/toxsci/kfs262. PubMed PMID: 22956630; PMCID: PMC3621350. 356. Quiros-Alcala L, Buckley JP, Boyle M. Parabens and measures of adiposity among adults and children from the U.S. general population: NHANES 2007-2014. Int J Hyg Environ Health. 2018;221(4):652-60. doi: 10.1016/j.ijheh.2018.03.006. PubMed PMID: 29580847; PMCID: PMC6685531. 357. Towers CV, Terry PD, Lewis D, Howard B, Chambers W, Armistead C, Weitz B, Porter S, Borman CJ, Kennedy RC, Chen J. Transplacental passage of antimicrobial paraben preservatives. J Expo Sci Environ Epidemiol. 2015;25(6):604-7. doi: 10.1038/jes.2015.27. PubMed PMID: 25944699. 358. Valle-Sistac J, Molins-Delgado D, Diaz M, Ibanez L, Barcelo D, Silvia Diaz-Cruz M. Determination of parabens and benzophenone-type UV filters in human placenta. First description of the existence of benzyl paraben and benzophenone-4. Environ Int. 2016;88:243-9. doi: 10.1016/j.envint.2015.12.034. PubMed PMID: 26773395. 359. Wu C, Xia W, Li Y, Li J, Zhang B, Zheng T, Zhou A, Zhao H, Huo W, Hu J, Jiang M, Hu C, Liao J, Chen X, Xu B, Lu S, Cai Z, Xu S. Repeated Measurements of Paraben Exposure during Pregnancy in Relation to Fetal and Early Childhood Growth. Environ Sci Technol. 2019;53(1):422-33. Epub 2018/11/15. doi: 10.1021/acs.est.8b01857. PubMed PMID: 30427191. 360. Chang CH, Wang PW, Liang HW, Huang YF, Huang LW, Chen HC, Pan WC, Lin MH, Yang W, Mao IF, Chen ML. The sex-specific association between maternal paraben exposure and size at birth. Int J Hyg Environ Health. 2019;222(6):955-64. Epub 2019/06/30. doi: 10.1016/j.ijheh.2019.06.004. PubMed PMID: 31248753. 361. Hajizadeh Y, Moradnia M, Kiani Feizabadi G, Rafiei N, Tahmasbizadeh M, Darvishmotevalli M, Fadaei S, Karimi H. The sex-specific association between maternal urinary paraben levels and offspring size at birth. Environ Sci Pollut Res Int. 2021;28(27):36029-38. doi: 10.1007/s11356-021-13175-3. PubMed PMID: 33683593. 362. Agier L, Basagana X, Hernandez-Ferrer C, Maitre L, Tamayo Uria I, Urquiza J, Andrusaityte S, Casas M, de Castro M, Cequier E, Chatzi L, Donaire-Gonzalez D, Giorgis-Allemand L, Gonzalez JR, Grazuleviciene R, Gutzkow KB, Haug LS, Sakhi AK, McEachan RRC, Meltzer HM, Nieuwenhuijsen M, Robinson O, Roumeliotaki T, Sunyer J, Thomsen C, Vafeiadi M, Valentin A, West J, Wright J, Siroux V, Vrijheid M, Slama R. Association between the pregnancy exposome and fetal growth. Int J Epidemiol. 2020;49(2):572-86. doi: 10.1093/ije/dyaa017. PubMed PMID: 32167557; PMCID: PMC7266545. 363. Aung MT, Ferguson KK, Cantonwine DE, McElrath TF, Meeker JD. Preterm birth in relation to the bisphenol A replacement, bisphenol S, and other phenols and parabens. Environ Res. 2019;169:131-8. Epub 2018/11/19. doi: 10.1016/j.envres.2018.10.037. PubMed PMID: 30448626; PMCID: PMC6347500. 185 364. Mustieles V, Zhang Y, Yland J, Braun JM, Williams PL, Wylie BJ, Attaman JA, Ford JB, Azevedo A, Calafat AM, Hauser R, Messerlian C. Maternal and paternal preconception exposure to phenols and preterm birth. Environment International. 2020;137:105523. doi: https://doi.org/10.1016/j.envint.2020.105523. 365. Jamal A, Rastkari N, Dehghaniathar R, Aghaei M, Nodehi RN, Nasseri S, Kashani H, Yunesian M. Prenatal exposure to parabens and anthropometric birth outcomes: A systematic review. Environ Res. 2019;173:419-31. doi: 10.1016/j.envres.2019.02.044. PubMed PMID: 30974368. 366. Zhong Q, Peng M, He J, Yang W, Huang F. Association of prenatal exposure to phenols and parabens with birth size: A systematic review and meta-analysis. Sci Total Environ. 2020;703:134720. doi: 10.1016/j.scitotenv.2019.134720. PubMed PMID: 31731171. 367. Vrijens K, Van Overmeire I, De Cremer K, Neven KY, Carollo RM, Vleminckx C, Van Loco J, Nawrot TS. Weight and head circumference at birth in function of placental paraben load in Belgium: an ENVIRONAGE birth cohort study. Environ Health. 2020;19(1):83. doi: 10.1186/s12940-020-00635-5. PubMed PMID: 32664952; PMCID: PMC7359508. 368. Jamal A, Rastkari N, Dehghaniathar R, Nodehi RN, Nasseri S, Kashani H, Shamsipour M, Yunesian M. Prenatal urinary concentrations of environmental phenols and birth outcomes in the mother-infant pairs of Tehran Environment and Neurodevelopmental Disorders (TEND) cohort study. Environmental Research. 2020;184:109331. doi: https://doi.org/10.1016/j.envres.2020.109331. 369. Chen LW, Aubert AM, Shivappa N, Bernard JY, Mensink-Bout SM, Geraghty AA, Mehegan J, Suderman M, Polanska K, Hanke W, Trafalska E, Relton CL, Crozier SR, Harvey NC, Cooper C, Duijts L, Heude B, Hebert JR, McAuliffe FM, Kelleher CC, Phillips CM. Associations of maternal dietary inflammatory potential and quality with offspring birth outcomes: An individual participant data pooled analysis of 7 European cohorts in the ALPHABET consortium. PLoS Med. 2021;18(1):e1003491. doi: 10.1371/journal.pmed.1003491. PubMed PMID: 33476335; PMCID: PMC7819611 following competing interests: JRH owns controlling interest in Connecting Health Innovations LLC (CHI), a company that has licensed the right to his invention of the dietary inflammatory index (DII) from the University of South Carolina in order to develop computer and smart phone applications for patient counselling and dietary intervention in clinical settings. NS is an employee of CHI. All other authors declare no support from any organisation for the submitted work other than those described above; no financial relationships with any organisations that might have an interest in the submitted work in the previous three years; no other relationships or activities that could appear to have influenced the submitted work. 370. Dolinoy DC, Huang D, Jirtle RL. Maternal nutrient supplementation counteracts bisphenol A-induced DNA hypomethylation in early development. Proc Natl Acad Sci U 186 S A. 2007;104(32):13056-61. Epub 2007/08/03. doi: 10.1073/pnas.0703739104. PubMed PMID: 17670942; PMCID: PMC1941790. 371. Oken E, Radesky JS, Wright RO, Bellinger DC, Amarasiriwardena CJ, Kleinman KP, Hu H, Gillman MW. Maternal fish intake during pregnancy, blood mercury levels, and child cognition at age 3 years in a US cohort. Am J Epidemiol. 2008;167(10):1171- 81. doi: 10.1093/aje/kwn034. PubMed PMID: 18353804; PMCID: PMC2590872. 372. Hertz-Picciotto I, Schramm M, Watt-Morse M, Chantala K, Anderson J, Osterloh J. Patterns and determinants of blood lead during pregnancy. Am J Epidemiol. 2000;152(9):829-37. doi: 10.1093/aje/152.9.829. PubMed PMID: 11085394. 373. Gennings C, Wolk A, Hakansson N, Lindh C, Bornehag CG. Contrasting prenatal nutrition and environmental exposures in association with birth weight and cognitive function in children at 7 years. BMJ Nutr Prev Health. 2020;3(2):162-71. doi: 10.1136/bmjnph-2020-000099. PubMed PMID: 33521525; PMCID: PMC7841844. 374. Zhang X, Chiu YM, Kannan S, Cowell W, Deng W, Coull BA, Wright RO, Wright RJ. Joint associations among prenatal metal mixtures and nutritional factors on birth weight z-score: Evidence from an urban U.S. population. Environ Res. 2022;208:112675. Epub 2022/01/08. doi: 10.1016/j.envres.2022.112675. PubMed PMID: 34995543; PMCID: PMC8916990. 375. Wen Q, Zhou Y, Wang Y, Li J, Zhao H, Liao J, Liu H, Li Y, Cai Z, Xia W. Association between urinary paraben concentrations and gestational weight gain during pregnancy. Journal of exposure science & environmental epidemiology. 2020;30(5):845-55. Epub 2020/02/12. doi: 10.1038/s41370-020-0205-7. PubMed PMID: 32042059. 376. Gonzalez-Nahm S, Nihlani K, J SH, R LM, H GS, Hoyo C. Associations between Maternal Cadmium Exposure with Risk of Preterm Birth and Low after Birth Weight Effect of Mediterranean Diet Adherence on Affected Prenatal Outcomes. Toxics. 2020;8(4). Epub 2020/10/24. doi: 10.3390/toxics8040090. PubMed PMID: 33092103; PMCID: PMC7712046. 377. Mahmassani HA, Switkowski KM, Scott TM, Johnson EJ, Rifas-Shiman SL, Oken E, Jacques PF. Maternal diet quality during pregnancy and child cognition and behavior in a US cohort. Am J Clin Nutr. 2021. Epub 2021/09/26. doi: 10.1093/ajcn/nqab325. PubMed PMID: 34562095. 378. Li M, Grewal J, Hinkle SN, Yisahak SF, Grobman WA, Newman RB, Skupski DW, Chien EK, Wing DA, Grantz KL, Zhang C. Healthy dietary patterns and common pregnancy complications: a prospective and longitudinal study. Am J Clin Nutr. 2021;114(3):1229-37. Epub 2021/06/03. doi: 10.1093/ajcn/nqab145. PubMed PMID: 34075392; PMCID: PMC8408886. 379. Reyes-Lopez MA, Gonzalez-Leyva CP, Rodriguez-Cano AM, Rodriguez- Hernandez C, Colin-Ramirez E, Estrada-Gutierrez G, Munoz-Manrique CG, Perichart- 187 Perera O. Diet Quality Is Associated with a High Newborn Size and Reduction in the Risk of Low Birth Weight and Small for Gestational Age in a Group of Mexican Pregnant Women: An Observational Study. Nutrients. 2021;13(6). Epub 2021/06/03. doi: 10.3390/nu13061853. PubMed PMID: 34071717; PMCID: PMC8227044. 380. Yisahak SF, Mumford SL, Grewal J, Li M, Zhang C, Grantz KL, Hinkle SN. Maternal diet patterns during early pregnancy in relation to neonatal outcomes. Am J Clin Nutr. 2021;114(1):358-67. Epub 2021/03/21. doi: 10.1093/ajcn/nqab019. PubMed PMID: 33742192; PMCID: PMC8246623. 381. Wirth MD, Hebert JR, Shivappa N, Hand GA, Hurley TG, Drenowatz C, McMahon D, Shook RP, Blair SN. Anti-inflammatory Dietary Inflammatory Index scores are associated with healthier scores on other dietary indices. Nutr Res. 2016;36(3):214- 9. doi: 10.1016/j.nutres.2015.11.009. PubMed PMID: 26923507; PMCID: PMC4773655. 382. Cox JL, Holden JM, Sagovsky R. Detection of postnatal depression. Development of the 10-item Edinburgh Postnatal Depression Scale. Br J Psychiatry. 1987;150:782-6. Epub 1987/06/01. doi: 10.1192/bjp.150.6.782. PubMed PMID: 3651732. 383. Ye X, Bishop AM, Reidy JA, Needham LL, Calafat AM. Parabens as urinary biomarkers of exposure in humans. Environ Health Perspect. 2006;114(12):1843-6. doi: 10.1289/ehp.9413. PubMed PMID: 17185273; PMCID: PMC1764178. 384. Ye X, Kuklenyik Z, Bishop AM, Needham LL, Calafat AM. Quantification of the urinary concentrations of parabens in humans by on-line solid phase extraction-high performance liquid chromatography-isotope dilution tandem mass spectrometry. J Chromatogr B Analyt Technol Biomed Life Sci. 2006;844(1):53-9. doi: 10.1016/j.jchromb.2006.06.037. PubMed PMID: 16893688. 385. Talge NM, Mudd LM, Sikorskii A, Basso O. United States birth weight reference corrected for implausible gestational age estimates. Pediatrics. 2014;133(5):844-53. doi: 10.1542/peds.2013-3285. PubMed PMID: 24777216. 386. Committee opinion no. 633: Alcohol abuse and other substance use disorders: ethical issues in obstetric and gynecologic practice. Obstet Gynecol. 2015;125(6):1529- 37. doi: 10.1097/01.AOG.0000466371.86393.9b. PubMed PMID: 26000541. 387. Emond JA, Karagas MR, Baker ER, Gilbert-Diamond D. Better Diet Quality during Pregnancy Is Associated with a Reduced Likelihood of an Infant Born Small for Gestational Age: An Analysis of the Prospective New Hampshire Birth Cohort Study. J Nutr. 2018;148(1):22-30. doi: 10.1093/jn/nxx005. PubMed PMID: 29378041; PMCID: PMC6251578. 388. Kamai EM, McElrath TF, Ferguson KK. Fetal growth in environmental epidemiology: mechanisms, limitations, and a review of associations with biomarkers of non-persistent chemical exposures during pregnancy. Environ Health. 2019;18(1):43. 188 Epub 2019/05/10. doi: 10.1186/s12940-019-0480-8. PubMed PMID: 31068204; PMCID: PMC6505101. 389. Risnes KR, Vatten LJ, Baker JL, Jameson K, Sovio U, Kajantie E, Osler M, Morley R, Jokela M, Painter RC, Sundh V, Jacobsen GW, Eriksson JG, Sorensen TI, Bracken MB. Birthweight and mortality in adulthood: a systematic review and meta- analysis. Int J Epidemiol. 2011;40(3):647-61. Epub 2011/02/18. doi: 10.1093/ije/dyq267. PubMed PMID: 21324938. 390. Cowell WJ, Wright RJ. Sex-Specific Effects of Combined Exposure to Chemical and Non-chemical Stressors on Neuroendocrine Development: a Review of Recent Findings and Putative Mechanisms. Curr Environ Health Rep. 2017;4(4):415-25. Epub 2017/10/14. doi: 10.1007/s40572-017-0165-9. PubMed PMID: 29027649; PMCID: PMC6416052. 391. Gabory A, Attig L, Junien C. Sexual dimorphism in environmental epigenetic programming. Mol Cell Endocrinol. 2009;304(1-2):8-18. Epub 2009/05/13. doi: 10.1016/j.mce.2009.02.015. PubMed PMID: 19433243. 392. Fadaei S, Pourzamani H, Ebrahimpour K, Feizi A, Daniali SS, Kelishadi R. Association of maternal urinary concentration of parabens and neonatal anthropometric indices. J Environ Health Sci Eng. 2020;18(2):617-28. Epub 2020/12/15. doi: 10.1007/s40201-020-00487-8. PubMed PMID: 33312588; PMCID: PMC7721758. 393. Philippat C, Botton J, Calafat AM, Ye X, Charles MA, Slama R, Group ES. Prenatal exposure to phenols and growth in boys. Epidemiology. 2014;25(5):625-35. Epub 2014/07/26. doi: 10.1097/EDE.0000000000000132. PubMed PMID: 25061923; PMCID: PMC4724208. 394. Wu C, Huo W, Li Y, Zhang B, Wan Y, Zheng T, Zhou A, Chen Z, Qian M, Zhu Y, Jiang Y, Liu H, Hu J, Chen X, Xu B, Xia W, Xu S. Maternal urinary paraben levels and offspring size at birth from a Chinese birth cohort. Chemosphere. 2017;172:29-36. Epub 2017/01/07. doi: 10.1016/j.chemosphere.2016.12.131. PubMed PMID: 28061343. 395. Geer LA, Pycke BFG, Waxenbaum J, Sherer DM, Abulafia O, Halden RU. Association of birth outcomes with fetal exposure to parabens, triclosan and triclocarban in an immigrant population in Brooklyn, New York. J Hazard Mater. 2017;323(Pt A):177- 83. Epub 2016/05/10. doi: 10.1016/j.jhazmat.2016.03.028. PubMed PMID: 27156397; PMCID: PMC5018415. 396. Messerlian C, Mustieles V, Minguez-Alarcon L, Ford JB, Calafat AM, Souter I, Williams PL, Hauser R, Environment, Reproductive Health Study T. Preconception and prenatal urinary concentrations of phenols and birth size of singleton infants born to mothers and fathers from the Environment and Reproductive Health (EARTH) study. Environ Int. 2018;114:60-8. Epub 2018/02/27. doi: 10.1016/j.envint.2018.02.017. PubMed PMID: 29477955; PMCID: PMC5899953. 189 397. Ferguson KK, Meeker JD, Cantonwine DE, Mukherjee B, Pace GG, Weller D, McElrath TF. Environmental phenol associations with ultrasound and delivery measures of fetal growth. Environ Int. 2018;112:243-50. Epub 2018/01/03. doi: 10.1016/j.envint.2017.12.011. PubMed PMID: 29294443; PMCID: PMC5899051. 398. Philippat C, Mortamais M, Chevrier C, Petit C, Calafat AM, Ye X, Silva MJ, Brambilla C, Pin I, Charles MA, Cordier S, Slama R. Exposure to phthalates and phenols during pregnancy and offspring size at birth. Environ Health Perspect. 2012;120(3):464-70. Epub 2011/09/09. doi: 10.1289/ehp.1103634. PubMed PMID: 21900077; PMCID: PMC3295340. 399. Aker AM, Ferguson KK, Rosario ZY, Mukherjee B, Alshawabkeh AN, Cordero JF, Meeker JD. The associations between prenatal exposure to triclocarban, phenols and parabens with gestational age and birth weight in northern Puerto Rico. Environ Res. 2019;169:41-51. Epub 2018/11/10. doi: 10.1016/j.envres.2018.10.030. PubMed PMID: 30412856; PMCID: PMC6347499. 400. Goodrich JM, Ingle ME, Domino SE, Treadwell MC, Dolinoy DC, Burant C, Meeker JD, Padmanabhan V. First trimester maternal exposures to endocrine disrupting chemicals and metals and fetal size in the Michigan Mother-Infant Pairs study. J Dev Orig Health Dis. 2019;10(4):447-58. Epub 2019/01/31. doi: 10.1017/S204017441800106X. PubMed PMID: 30696509; PMCID: PMC6660406. 401. Pycke BF, Geer LA, Dalloul M, Abulafia O, Halden RU. Maternal and fetal exposure to parabens in a multiethnic urban U.S. population. Environ Int. 2015;84:193- 200. doi: 10.1016/j.envint.2015.08.012. PubMed PMID: 26364793; PMCID: PMC4613774. 402. Brauner EV, Uldbjerg CS, Lim YH, Gregersen LS, Krause M, Frederiksen H, Andersson AM. Presence of parabens, phenols and phthalates in paired maternal serum, urine and amniotic fluid. Environ Int. 2022;158:106987. doi: 10.1016/j.envint.2021.106987. PubMed PMID: 34991249; PMCID: PMC8739868. 403. Vandentorren S, Zeman F, Morin L, Sarter H, Bidondo ML, Oleko A, Leridon H. Bisphenol-A and phthalates contamination of urine samples by catheters in the Elfe pilot study: implications for large-scale biomonitoring studies. Environ Res. 2011;111(6):761- 4. doi: 10.1016/j.envres.2011.05.018. PubMed PMID: 21684541. 404. Aker AM, Ferguson KK, Rosario ZY, Mukherjee B, Alshawabkeh AN, Calafat AM, Cordero JF, Meeker JD. A repeated measures study of phenol, paraben and Triclocarban urinary biomarkers and circulating maternal hormones during gestation in the Puerto Rico PROTECT cohort. Environ Health. 2019;18(1):28. Epub 2019/04/04. doi: 10.1186/s12940-019-0459-5. PubMed PMID: 30940137; PMCID: PMC6444601. 405. Rosen ED, Spiegelman BM. Adipocytes as regulators of energy balance and glucose homeostasis. Nature. 2006;444(7121):847-53. Epub 2006/12/15. doi: 10.1038/nature05483. PubMed PMID: 17167472; PMCID: PMC3212857. 190 406. Bellavia A, Chiu YH, Brown FM, Minguez-Alarcon L, Ford JB, Keller M, Petrozza J, Williams PL, Ye X, Calafat AM, Hauser R, James-Todd T, Team ES. Urinary concentrations of parabens mixture and pregnancy glucose levels among women from a fertility clinic. Environ Res. 2019;168:389-96. Epub 2018/11/02. doi: 10.1016/j.envres.2018.10.009. PubMed PMID: 30384233; PMCID: PMC7190006. 407. Al-Qaraghouli M, Fang YMV. Effect of Fetal Sex on Maternal and Obstetric Outcomes. Front Pediatr. 2017;5:144. Epub 2017/07/05. doi: 10.3389/fped.2017.00144. PubMed PMID: 28674684; PMCID: PMC5476168. 408. Enninga EA, Nevala WK, Creedon DJ, Markovic SN, Holtan SG. Fetal sex-based differences in maternal hormones, angiogenic factors, and immune mediators during pregnancy and the postpartum period. Am J Reprod Immunol. 2015;73(3):251-62. Epub 2014/08/06. doi: 10.1111/aji.12303. PubMed PMID: 25091957; PMCID: PMC4317383. 409. Rosenfeld CS. Sex-Specific Placental Responses in Fetal Development. Endocrinology. 2015;156(10):3422-34. Epub 2015/08/05. doi: 10.1210/en.2015-1227. PubMed PMID: 26241064; PMCID: PMC4588817. 410. Drife JO. The history of labour induction: How did we get here? Best Pract Res Clin Obstet Gynaecol. 2021. Epub 2021/08/01. doi: 10.1016/j.bpobgyn.2021.07.004. PubMed PMID: 34330639. 411. Zhang Y, Mustieles V, Williams PL, Yland J, Souter I, Braun JM, Calafat AM, Hauser R, Messerlian C. Prenatal urinary concentrations of phenols and risk of preterm birth: exploring windows of vulnerability. Fertil Steril. 2021;116(3):820-32. Epub 2021/07/10. doi: 10.1016/j.fertnstert.2021.03.053. PubMed PMID: 34238571. 412. Vernet C, Philippat C, Calafat AM, Ye X, Lyon-Caen S, Siroux V, Schisterman EF, Slama R. Within-Day, Between-Day, and Between-Week Variability of Urinary Concentrations of Phenol Biomarkers in Pregnant Women. Environmental Health Perspectives. 2018;126(3):037005. Epub 2018/03/20. doi: 10.1289/ehp1994. PubMed PMID: 29553460; PMCID: PMC6071804. 413. Vargas-Terrones M, Nagpal TS, Barakat R. Impact of exercise during pregnancy on gestational weight gain and birth weight: an overview. Braz J Phys Ther. 2019;23(2):164-9. Epub 2018/12/12. doi: 10.1016/j.bjpt.2018.11.012. PubMed PMID: 30527949; PMCID: PMC6428912. 414. Fadaei S, Pourzamani H, Ebrahimpour K, Feizi A, Daniali SS, Kelishadi R. Investigating determinants of parabens concentration in maternal urine. Hum Ecol Risk Assess. 2021;27(3):668-86. doi: 10.1080/10807039.2020.1750344. PubMed PMID: WOS:000544167100001. 415. Bilodeau JF, Bisson M, Larose J, Pronovost E, Brien M, Greffard K, Marc I. Physical fitness is associated with prostaglandin F2alpha isomers during pregnancy. Prostaglandins Leukot Essent Fatty Acids. 2019;145:7-14. Epub 2019/05/28. doi: 10.1016/j.plefa.2019.05.001. PubMed PMID: 31126516. 191 416. Mudd LM, Evenson KR. Review of impacts of physical activity on maternal metabolic health during pregnancy. Curr Diab Rep. 2015;15(2):572. Epub 2015/01/31. doi: 10.1007/s11892-014-0572-3. PubMed PMID: 25633442. 417. Shin HM, Oh J, Kim K, Busgang SA, Barr DB, Panuwet P, Schmidt RJ, Hertz- Picciotto I, Bennett DH. Variability of Urinary Concentrations of Phenols, Parabens, and Triclocarban during Pregnancy in First Morning Voids and Pooled Samples. Environ Sci Technol. 2021;55(23):16001-10. doi: 10.1021/acs.est.1c04140. PubMed PMID: 34817155; PMCID: PMC8858442. 418. Mou L, Norby FL, Chen LY, O'Neal WT, Lewis TT, Loehr LR, Soliman EZ, Alonso A. Lifetime Risk of Atrial Fibrillation by Race and Socioeconomic Status: ARIC Study (Atherosclerosis Risk in Communities). Circ Arrhythm Electrophysiol. 2018;11(7):e006350. Epub 2018/07/14. doi: 10.1161/CIRCEP.118.006350. PubMed PMID: 30002066; PMCID: PMC6053683. 419. Organization WH. Obesity and Overweight 2018 [cited 2019 March 13]. Available from: https://www.who.int/news-room/fact-sheets/detail/obesity-and-overweight. 420. Hales CM, Carroll MD, Fryar CD, Ogden CL. Prevalence of Obesity Among Adults and Youth: United States, 2015-2016. NCHS Data Brief. 2017(288):1-8. Epub 2017/11/21. PubMed PMID: 29155689. 421. Shai I, Jiang R, Manson JE, Stampfer MJ, Willett WC, Colditz GA, Hu FB. Ethnicity, obesity, and risk of type 2 diabetes in women: a 20-year follow-up study. Diabetes Care. 2006;29(7):1585-90. Epub 2006/06/28. doi: 10.2337/dc06-0057. PubMed PMID: 16801583. 422. Antillon D, Towfighi A. No time to 'weight': the link between obesity and stroke in women. Womens Health (Lond). 2011;7(4):453-63. Epub 2011/07/28. doi: 10.2217/whe.11.36. PubMed PMID: 21790338. 423. Mulugeta A, Zhou A, Power C, Hypponen E. Obesity and depressive symptoms in mid-life: a population-based cohort study. BMC Psychiatry. 2018;18(1):297. Epub 2018/09/22. doi: 10.1186/s12888-018-1877-6. PubMed PMID: 30236085; PMCID: PMC6148790. 424. Gelber RP, Gaziano JM, Orav EJ, Manson JE, Buring JE, Kurth T. Measures of obesity and cardiovascular risk among men and women. J Am Coll Cardiol. 2008;52(8):605-15. Epub 2008/08/16. doi: 10.1016/j.jacc.2008.03.066. PubMed PMID: 18702962; PMCID: PMC2671389. 425. Singh-Manoux A, Dugravot A, Shipley M, Brunner EJ, Elbaz A, Sabia S, Kivimaki M. Obesity trajectories and risk of dementia: 28 years of follow-up in the Whitehall II Study. Alzheimers Dement. 2018;14(2):178-86. Epub 2017/09/26. doi: 10.1016/j.jalz.2017.06.2637. PubMed PMID: 28943197; PMCID: PMC5805839. 192 426. Lauby-Secretan B, Scoccianti C, Loomis D, Grosse Y, Bianchini F, Straif K, International Agency for Research on Cancer Handbook Working G. Body Fatness and Cancer--Viewpoint of the IARC Working Group. N Engl J Med. 2016;375(8):794-8. Epub 2016/08/25. doi: 10.1056/NEJMsr1606602. PubMed PMID: 27557308. 427. Morabia A, Costanza MC. International variability in ages at menarche, first livebirth, and menopause. World Health Organization Collaborative Study of Neoplasia and Steroid Contraceptives. Am J Epidemiol. 1998;148(12):1195-205. Epub 1998/12/29. PubMed PMID: 9867266. 428. Santoro N, Randolph JF, Jr. Reproductive hormones and the menopause transition. Obstet Gynecol Clin North Am. 2011;38(3):455-66. Epub 2011/10/04. doi: 10.1016/j.ogc.2011.05.004. PubMed PMID: 21961713; PMCID: PMC3197715. 429. Bacon JL. The Menopausal Transition. Obstet Gynecol Clin North Am. 2017;44(2):285-96. Epub 2017/05/14. doi: 10.1016/j.ogc.2017.02.008. PubMed PMID: 28499537. 430. Lovejoy JC. The influence of sex hormones on obesity across the female life span. J Womens Health. 1998;7(10):1247-56. Epub 1999/02/04. doi: 10.1089/jwh.1998.7.1247. PubMed PMID: 9929857. 431. Toth MJ, Tchernof A, Sites CK, Poehlman ET. Effect of menopausal status on body composition and abdominal fat distribution. Int J Obes Relat Metab Disord. 2000;24(2):226-31. Epub 2000/03/07. PubMed PMID: 10702775. 432. Lumsden MAHK. Impact of obesity on the health of women in midlife. The Obstetrician & Gynaecologist. 2015;17(201):8. 433. Sutton-Tyrrell K, Wildman RP, Matthews KA, Chae C, Lasley BL, Brockwell S, Pasternak RC, Lloyd-Jones D, Sowers MF, Torrens JI, Investigators S. Sex-hormone- binding globulin and the free androgen index are related to cardiovascular risk factors in multiethnic premenopausal and perimenopausal women enrolled in the Study of Women Across the Nation (SWAN). Circulation. 2005;111(10):1242-9. Epub 2005/03/17. doi: 10.1161/01.CIR.0000157697.54255.CE. PubMed PMID: 15769764. 434. Sutton-Tyrrell K, Zhao X, Santoro N, Lasley B, Sowers M, Johnston J, Mackey R, Matthews K. Reproductive hormones and obesity: 9 years of observation from the Study of Women's Health Across the Nation. Am J Epidemiol. 2010;171(11):1203-13. Epub 2010/04/30. doi: 10.1093/aje/kwq049. PubMed PMID: 20427327; PMCID: PMC2915490. 435. Tao X, Jiang A, Yin L, Li Y, Tao F, Hu H. Body mass index and age at natural menopause: a meta-analysis. Menopause. 2015;22(4):469-74. Epub 2014/09/10. doi: 10.1097/GME.0000000000000324. PubMed PMID: 25203893. 436. Ley SH, Li Y, Tobias DK, Manson JE, Rosner B, Hu FB, Rexrode KM. Duration of Reproductive Life Span, Age at Menarche, and Age at Menopause Are Associated 193 With Risk of Cardiovascular Disease in Women. J Am Heart Assoc. 2017;6(11). Epub 2017/11/04. doi: 10.1161/JAHA.117.006713. PubMed PMID: 29097389; PMCID: PMC5721766. 437. Collaborative Group on Hormonal Factors in Breast C. Menarche, menopause, and breast cancer risk: individual participant meta-analysis, including 118 964 women with breast cancer from 117 epidemiological studies. Lancet Oncol. 2012;13(11):1141- 51. Epub 2012/10/23. doi: 10.1016/S1470-2045(12)70425-4. PubMed PMID: 23084519; PMCID: PMC3488186. 438. Gold EB, Sternfeld B, Kelsey JL, Brown C, Mouton C, Reame N, Salamone L, Stellato R. Relation of demographic and lifestyle factors to symptoms in a multi- racial/ethnic population of women 40-55 years of age. Am J Epidemiol. 2000;152(5):463-73. Epub 2000/09/12. PubMed PMID: 10981461. 439. Pastore LM, Carter RA, Hulka BS, Wells E. Self-reported urogenital symptoms in postmenopausal women: Women's Health Initiative. Maturitas. 2004;49(4):292-303. Epub 2004/11/09. doi: 10.1016/j.maturitas.2004.06.019. PubMed PMID: 15531125. 440. Schilling C, Gallicchio L, Miller SR, Langenberg P, Zacur H, Flaws JA. Relation of body mass and sex steroid hormone levels to hot flushes in a sample of mid-life women. Climacteric. 2007;10(1):27-37. Epub 2007/03/17. doi: 10.1080/13697130601164755. PubMed PMID: 17364602. 441. We JS, Han K, Kwon HS, Kil K. Effect of Maternal Age at Childbirth on Obesity in Postmenopausal Women: A Nationwide Population-Based Study in Korea. Medicine (Baltimore). 2016;95(19):e3584. Epub 2016/05/14. doi: 10.1097/MD.0000000000003584. PubMed PMID: 27175656; PMCID: PMC4902498. 442. Weng HH, Bastian LA, Taylor DH, Jr., Moser BK, Ostbye T. Number of children associated with obesity in middle-aged women and men: results from the health and retirement study. J Womens Health (Larchmt). 2004;13(1):85-91. Epub 2004/03/10. doi: 10.1089/154099904322836492. PubMed PMID: 15006281. 443. Franca AP, Marucci MFN, Silva M, Roediger MA. [Factors associated with general obesity and the percentage of body fat of women during the menopause in the city of Sao Paulo, Brazil]. Cien Saude Colet. 2018;23(11):3577-86. Epub 2018/11/15. doi: 10.1590/1413-812320182311.26492016. PubMed PMID: 30427431. 444. Lao XQ, Thomas GN, Jiang CQ, Zhang WS, Yin P, Schooling M, Heys M, Leung GM, Adab P, Cheng KK, Lam TH. Parity and the metabolic syndrome in older Chinese women: the Guangzhou Biobank Cohort Study. Clin Endocrinol (Oxf). 2006;65(4):460-9. Epub 2006/09/21. doi: 10.1111/j.1365-2265.2006.02615.x. PubMed PMID: 16984238. 445. Hajiahmadi M, Shafi H, Delavar MA. Impact of parity on obesity: a cross- sectional study in Iranian women. Med Princ Pract. 2015;24(1):70-4. Epub 2014/11/18. doi: 10.1159/000368358. PubMed PMID: 25402350; PMCID: PMC5588186. 194 446. Martinez ME, Pond E, Wertheim BC, Nodora JN, Jacobs ET, Bondy M, Daneri- Navarro A, Meza-Montenegro MM, Gutierrez-Millan LE, Brewster A, Komenaka IK, Thompson P. Association between parity and obesity in Mexican and Mexican- American women: findings from the Ella binational breast cancer study. J Immigr Minor Health. 2013;15(2):234-43. Epub 2012/05/24. doi: 10.1007/s10903-012-9649-8. PubMed PMID: 22618357; PMCID: PMC3469728. 447. Trikudanathan S, Pedley A, Massaro JM, Hoffmann U, Seely EW, Murabito JM, Fox CS. Association of female reproductive factors with body composition: the Framingham Heart Study. J Clin Endocrinol Metab. 2013;98(1):236-44. Epub 2012/10/25. doi: 10.1210/jc.2012-1785. PubMed PMID: 23093491; PMCID: PMC3537091. 448. Kuijper EA, Ket JC, Caanen MR, Lambalk CB. Reproductive hormone concentrations in pregnancy and neonates: a systematic review. Reprod Biomed Online. 2013;27(1):33-63. Epub 2013/05/15. doi: 10.1016/j.rbmo.2013.03.009. PubMed PMID: 23669015. 449. Ziv-Gal A, Smith RL, Gallicchio L, Miller SR, Zacur HA, Flaws JA. The Midlife Women's Health Study - a study protocol of a longitudinal prospective study on predictors of menopausal hot flashes. Womens Midlife Health. 2017;3:4. Epub 2017/08/17. doi: 10.1186/s40695-017-0024-8. PubMed PMID: 30766705; PMCID: PMC6300019. 450. (CDC) CoDCaP. About Adult BMI 2017 [cited 2019 March 13]. Available from: https://www.cdc.gov/healthyweight/assessing/bmi/adult_bmi/index.html. 451. Smith RL, Gallicchio L, Flaws JA. Factors Affecting Sexual Activity in Midlife Women: Results from the Midlife Health Study. J Womens Health (Larchmt). 2017;26(2):103-8. Epub 2016/09/23. doi: 10.1089/jwh.2016.5881. PubMed PMID: 27653205; PMCID: PMC5312621. 452. Chiang C, Gallicchio L, Zacur H, Miller S, Flaws JA, Smith RL. Hormone variability and hot flash experience: Results from the midlife women's health study. Maturitas. 2019;119:1-7. Epub 2018/12/07. doi: 10.1016/j.maturitas.2018.10.007. PubMed PMID: 30502745; PMCID: PMC6289582. 453. Greenland S, Pearl J, Robins JM. Causal diagrams for epidemiologic research. Epidemiology. 1999;10(1):37-48. Epub 1999/01/15. PubMed PMID: 9888278. 454. Pirkle CM, de Albuquerque Sousa AC, Alvarado B, Zunzunegui MV, Group IR. Early maternal age at first birth is associated with chronic diseases and poor physical performance in older age: cross-sectional analysis from the International Mobility in Aging Study. BMC Public Health. 2014;14:293. Epub 2014/04/02. doi: 10.1186/1471- 2458-14-293. PubMed PMID: 24684705; PMCID: PMC3977880. 195 455. Richiardi L, Bellocco R, Zugna D. Mediation analysis in epidemiology: methods, interpretation and bias. Int J Epidemiol. 2013;42(5):1511-9. Epub 2013/09/11. doi: 10.1093/ije/dyt127. PubMed PMID: 24019424. 456. Kinoshita T, Itoh M. Longitudinal variance of fat mass deposition during pregnancy evaluated by ultrasonography: the ratio of visceral fat to subcutaneous fat in the abdomen. Gynecol Obstet Invest. 2006;61(2):115-8. Epub 2005/11/08. doi: 10.1159/000089456. PubMed PMID: 16272815. 457. To WW, Wong MW. Body fat composition and weight changes during pregnancy and 6-8 months post-partum in primiparous and multiparous women. Aust N Z J Obstet Gynaecol. 2009;49(1):34-8. Epub 2009/03/14. doi: 10.1111/j.1479-828X.2008.00933.x. PubMed PMID: 19281577. 458. Lan-Pidhainy X, Nohr EA, Rasmussen KM. Comparison of gestational weight gain-related pregnancy outcomes in American primiparous and multiparous women. Am J Clin Nutr. 2013;97(5):1100-6. Epub 2013/04/05. doi: 10.3945/ajcn.112.052258. PubMed PMID: 23553161. 459. Martinez-Galiano JM, Hernandez-Martinez A, Rodriguez-Almagro J, Delgado- Rodriguez M, Gomez-Salgado J. Relationship between parity and the problems that appear in the postpartum period. Sci Rep. 2019;9(1):11763. Epub 2019/08/15. doi: 10.1038/s41598-019-47881-3. PubMed PMID: 31409871; PMCID: PMC6692385. 460. Bastian LA, Pathiraja VC, Krause K, Namenek Brouwer RJ, Swamy GK, Lovelady CA, Ostbye T. Multiparity is associated with high motivation to change diet among overweight and obese postpartum women. Womens Health Issues. 2010;20(2):133-8. Epub 2010/02/13. doi: 10.1016/j.whi.2009.11.005. PubMed PMID: 20149971; PMCID: PMC2849268. 461. Smith DE, Lewis CE, Caveny JL, Perkins LL, Burke GL, Bild DE. Longitudinal changes in adiposity associated with pregnancy. The CARDIA Study. Coronary Artery Risk Development in Young Adults Study. JAMA. 1994;271(22):1747-51. Epub 1994/06/08. PubMed PMID: 8196117. 462. Gunderson EP, Murtaugh MA, Lewis CE, Quesenberry CP, West DS, Sidney S. Excess gains in weight and waist circumference associated with childbearing: The Coronary Artery Risk Development in Young Adults Study (CARDIA). Int J Obes Relat Metab Disord. 2004;28(4):525-35. Epub 2004/02/11. doi: 10.1038/sj.ijo.0802551. PubMed PMID: 14770188; PMCID: PMC3133634. 463. Huayanay-Espinoza CA, Quispe R, Poterico JA, Carrillo-Larco RM, Bazo-Alvarez JC, Miranda JJ. Parity and Overweight/Obesity in Peruvian Women. Prev Chronic Dis. 2017;14:E102. Epub 2017/10/27. doi: 10.5888/pcd14.160282. PubMed PMID: 29072986; PMCID: PMC5662294. 464. Koch E, Bogado M, Araya F, Romero T, Diaz C, Manriquez L, Paredes M, Roman C, Taylor A, Kirschbaum A. Impact of parity on anthropometric measures of 196 obesity controlling by multiple confounders: a cross-sectional study in Chilean women. J Epidemiol Community Health. 2008;62(5):461-70. Epub 2008/04/17. doi: 10.1136/jech.2007.062240. PubMed PMID: 18413461. 465. Kim JH, Kim J, Ahn HJ, Kim SY, Bae HY. Impact of Parity on Body Size Phenotype in Postmenopausal Women: KNHANES 2010-2012. J Clin Endocrinol Metab. 2016;101(12):4904-13. Epub 2016/09/28. doi: 10.1210/jc.2016-2823. PubMed PMID: 27676397. 466. Gravena AA, Brischiliari SC, Lopes TC, Agnolo CM, Carvalho MD, Pelloso SM. Excess weight and abdominal obesity in postmenopausal Brazilian women: a population-based study. BMC Womens Health. 2013;13:46. Epub 2013/11/16. doi: 10.1186/1472-6874-13-46. PubMed PMID: 24228934; PMCID: PMC3833652. 467. Nelson SM, Matthews P, Poston L. Maternal metabolism and obesity: modifiable determinants of pregnancy outcome. Hum Reprod Update. 2010;16(3):255-75. Epub 2009/12/08. doi: 10.1093/humupd/dmp050. PubMed PMID: 19966268; PMCID: PMC2849703. 468. Kaaja RJ, Greer IA. Manifestations of chronic disease during pregnancy. JAMA. 2005;294(21):2751-7. Epub 2005/12/08. doi: 10.1001/jama.294.21.2751. PubMed PMID: 16333011. 469. Camara SM, Pirkle C, Moreira MA, Vieira MC, Vafaei A, Maciel AC. Early maternal age and multiparity are associated to poor physical performance in middle- aged women from Northeast Brazil: a cross-sectional community based study. BMC Womens Health. 2015;15:56. Epub 2015/08/06. doi: 10.1186/s12905-015-0214-1. PubMed PMID: 26243283; PMCID: PMC4526418. 470. Yang L, Li L, Millwood IY, Lewington S, Guo Y, Sherliker P, Peters SA, Bian Z, Wu X, Yu M, Liu H, Wang H, Mao E, Chen J, Woodward M, Peto R, Chen Z, China Kadoorie Biobank study collaborative g. Adiposity in relation to age at menarche and other reproductive factors among 300 000 Chinese women: findings from China Kadoorie Biobank study. Int J Epidemiol. 2017;46(2):502-12. Epub 2016/08/16. doi: 10.1093/ije/dyw165. PubMed PMID: 27524817; PMCID: PMC5837303. 471. Patchen L, Leoutsakos JM, Astone NM. Early Parturition: Is Young Maternal Age at First Birth Associated with Obesity? J Pediatr Adolesc Gynecol. 2017;30(5):553-9. Epub 2016/12/25. doi: 10.1016/j.jpag.2016.12.001. PubMed PMID: 28011235. 472. Gomez-Scott J, Cooney TM. Young women's education and behavioural risk trajectories: clarifying their association with unintended-pregnancy resolution. Cult Health Sex. 2014;16(6):648-65. Epub 2014/04/17. doi: 10.1080/13691058.2014.898794. PubMed PMID: 24735279. 473. Rifas-Shiman SL, Rich-Edwards JW, Kleinman KP, Oken E, Gillman MW. Dietary quality during pregnancy varies by maternal characteristics in Project Viva: a US cohort. 197 J Am Diet Assoc. 2009;109(6):1004-11. Epub 2009/05/26. doi: 10.1016/j.jada.2009.03.001. PubMed PMID: 19465182; PMCID: PMC4098830. 474. Kirchengast S, Gruber D, Sator M, Huber J. Postmenopausal weight status, body composition and body fat distribution in relation to parameters of menstrual and reproductive history. Maturitas. 1999;33(2):117-26. Epub 1999/12/22. PubMed PMID: 10597875. 475. Freeman EW, Sammel MD, Lin H, Gracia CR. Obesity and reproductive hormone levels in the transition to menopause. Menopause. 2010;17(4):718-26. Epub 2010/03/11. doi: 10.1097/gme.0b013e3181cec85d. PubMed PMID: 20216473; PMCID: PMC2888623. 476. Tilley BC, Barnes AB, Bergstralh E, Labarthe D, Noller KL, Colton T, Adam E. A comparison of pregnancy history recall and medical records. Implications for retrospective studies. Am J Epidemiol. 1985;121(2):269-81. Epub 1985/02/01. PubMed PMID: 3893098. 477. Liu J, Tuvblad C, Li L, Raine A, Baker LA. Medical record validation of maternal recall of pregnancy and birth events from a twin cohort. Twin Res Hum Genet. 2013;16(4):845-60. Epub 2013/06/04. doi: 10.1017/thg.2013.31. PubMed PMID: 23725849; PMCID: PMC4255463. 478. Dahl AK, Reynolds CA. Accuracy of recalled body weight--a study with 20-years of follow-up. Obesity (Silver Spring). 2013;21(6):1293-8. Epub 2013/08/06. doi: 10.1002/oby.20299. PubMed PMID: 23913738; PMCID: PMC3740460. 479. Benjamin S, Masai E, Kamimura N, Takahashi K, Anderson RC, Faisal PA. Phthalates impact human health: Epidemiological evidences and plausible mechanism of action. J Hazard Mater. 2017;340:360-83. Epub 06/19. doi: 10.1016/j.jhazmat.2017.06.036. PubMed PMID: 28800814. 480. Schettler T. Human exposure to phthalates via consumer products. Int J Androl. 2006;29(1):134-9; discussion 81-5. Epub 2006/02/10. doi: 10.1111/j.1365- 2605.2005.00567.x. PubMed PMID: 16466533. 481. Weiss JM, Gustafsson Å, Gerde P, Bergman Å, Lindh CH, Krais AM. Daily intake of phthalates, MEHP, and DINCH by ingestion and inhalation. Chemosphere. 2018;208:40-9. doi: https://doi.org/10.1016/j.chemosphere.2018.05.094. 482. Ding M, Kang Q, Zhang S, Zhao F, Mu D, Zhang H, Yang M, Hu J. Contribution of phthalates and phthalate monoesters from drinking water to daily intakes for the general population. Chemosphere. 2019;229:125-31. doi: https://doi.org/10.1016/j.chemosphere.2019.05.023. 483. Garcia Ibarra V, Rodriguez Bernaldo de Quiros A, Paseiro Losada P, Sendon R. Identification of intentionally and non-intentionally added substances in plastic packaging materials and their migration into food products. Anal Bioanal Chem. 198 2018;410(16):3789-803. Epub 2018/05/08. doi: 10.1007/s00216-018-1058-y. PubMed PMID: 29732500. 484. Rastkari N, Zare Jeddi M, Yunesian M, Ahmadkhaniha R. The Effect of Storage Time, Temperature and Type of Packaging on the Release of Phthalate Esters into Packed Acidic Liquids. Food Technol Biotechnol. 2017;55(4):562-9. doi: 10.17113/ftb.55.04.17.5128. PubMed PMID: 29540990. 485. Marie C, Hamlaoui S, Bernard L, Bourdeaux D, Sautou V, Lémery D, Vendittelli F, Sauvant-Rochat M-P. Exposure of hospitalised pregnant women to plasticizers contained in medical devices. BMC women's health. 2017;17(1):45-. doi: 10.1186/s12905-017-0398-7. PubMed PMID: 28637458. 486. Calafat AM, Needham LL, Silva MJ, Lambert G. Exposure to di-(2-ethylhexyl) phthalate among premature neonates in a neonatal intensive care unit. Pediatrics. 2004;113(5):e429-34. Epub 2004/05/04. doi: 10.1542/peds.113.5.e429. PubMed PMID: 15121985. 487. Martine B, Marie-Jeanne T, Cendrine D, Fabrice A, Marc C. Assessment of Adult Human Exposure to Phthalate Esters in the Urban Centre of Paris (France). Bulletin of Environmental Contamination and Toxicology. 2013;90(1):91-6. doi: 10.1007/s00128- 012-0859-5. 488. Craig ZR, Ziv-Gal A. Pretty Good or Pretty Bad? The Ovary and Chemicals in Personal Care Products. Toxicological Sciences. 2017;162(2):349-60. doi: 10.1093/toxsci/kfx285. 489. Meeker JD, Ferguson KK. Urinary phthalate metabolites are associated with decreased serum testosterone in men, women, and children from NHANES 2011-2012. J Clin Endocrinol Metab. 2014;99(11):4346-52. Epub 2014/08/14. doi: 10.1210/jc.2014- 2555. PubMed PMID: 25121464. 490. EPA US. Biomonitoring: Phthalates. America's Children and the Environment. 2017(3rd). 491. Woodruff TJ, Zota AR, Schwartz JM. Environmental chemicals in pregnant women in the United States: NHANES 2003-2004. Environmental health perspectives. 2011;119(6):878-85. Epub 2011/01/14. doi: 10.1289/ehp.1002727. PubMed PMID: 21233055. 492. Brehm E, Rattan S, Gao L, Flaws JA. Prenatal Exposure to Di(2-Ethylhexyl) Phthalate Causes Long-Term Transgenerational Effects on Female Reproduction in Mice. Endocrinology. 2018;159(2):795-809. Epub 2017/12/12. doi: 10.1210/en.2017- 03004. PubMed PMID: 29228129; PMCID: PMC5774227. 493. Chang WH, Wu MH, Pan HA, Guo PL, Lee CC. Semen quality and insulin-like factor 3: Associations with urinary and seminal levels of phthalate metabolites in adult 199 males. Chemosphere. 2017;173:594-602. Epub 2017/02/06. doi: 10.1016/j.chemosphere.2017.01.056. PubMed PMID: 28152410. 494. Chiang C, Flaws JA. Subchronic Exposure to Di(2-ethylhexyl) Phthalate and Diisononyl Phthalate During Adulthood Has Immediate and Long-Term Reproductive Consequences in Female Mice. Toxicol Sci. 2019;168(2):620-31. Epub 2019/01/17. doi: 10.1093/toxsci/kfz013. PubMed PMID: 30649530; PMCID: PMC6432868. 495. Hannon PR, Flaws JA. The effects of phthalates on the ovary. Front Endocrinol (Lausanne). 2015;6:8. Epub 2015/02/24. doi: 10.3389/fendo.2015.00008. PubMed PMID: 25699018; PMCID: PMC4313599. 496. Hoyer BB, Lenters V, Giwercman A, Jonsson BAG, Toft G, Hougaard KS, Bonde JPE, Specht IO. Impact of Di-2-Ethylhexyl Phthalate Metabolites on Male Reproductive Function: a Systematic Review of Human Evidence. Curr Environ Health Rep. 2018;5(1):20-33. Epub 2018/02/23. doi: 10.1007/s40572-018-0174-3. PubMed PMID: 29468520. 497. Cathey AL, Watkins D, Rosario ZY, Vélez C, Alshawabkeh AN, Cordero JF, Meeker JD. Associations of Phthalates and Phthalate Replacements With CRH and Other Hormones Among Pregnant Women in Puerto Rico. Journal of the Endocrine Society. 2019;3(6):1127-49. doi: 10.1210/js.2019-00010. PubMed PMID: 31093596. 498. Duty SM, Calafat AM, Silva MJ, Ryan L, Hauser R. Phthalate exposure and reproductive hormones in adult men. Human reproduction (Oxford, England). 2005;20(3):604-10. Epub 12/09. doi: 10.1093/humrep/deh656. PubMed PMID: 15591081. 499. Meeker JD, Calafat AM, Hauser R. Urinary Metabolites of Di(2-ethylhexyl) Phthalate Are Associated With Decreased Steroid Hormone Levels in Adult Men. Journal of Andrology. 2009;30(3):287-97. doi: 10.2164/jandrol.108.006403. 500. Long SE KLTL, Jacobson MH. Urinary phthalate metabolites and alternatives and serum sex steroid hormones among pre- and postmenopausal women from NHANES, 2013-16. Sci Total Environ. 2021(https://doi.org/10.1016/j.scitotenv.2020.144560). Epub 2021. 501. Diaz Santana MV, Hankinson SE, Bigelow C, Sturgeon SR, Zoeller RT, Tinker L, Manson JAE, Calafat AM, Meliker JR, Reeves KW. Urinary concentrations of phthalate biomarkers and weight change among postmenopausal women: a prospective cohort study. Environ Health. 2019;18(1):20. doi: 10.1186/s12940-019-0458-6. PubMed PMID: 30866962; PMCID: PMC6417117. 502. Ziv-Gal A, Gallicchio L, Chiang C, Ther SN, Miller SR, Zacur HA, Dills RL, Flaws JA. Phthalate metabolite levels and menopausal hot flashes in midlife women. Reproductive toxicology (Elmsford, NY). 2016;60:76-81. Epub 2016/02/13. doi: 10.1016/j.reprotox.2016.02.001. PubMed PMID: 26867866; PMCID: PMC4867120. 200 503. DeFlorio-Barker SA, Turyk ME. Associations between bone mineral density and urinary phthalate metabolites among post-menopausal women: a cross-sectional study of NHANES data 2005-2010. Int J Environ Health Res. 2016;26(3):326-45. Epub 2015/11/21. doi: 10.1080/09603123.2015.1111312. PubMed PMID: 26586408. 504. Gore AC, Hall JE, Hayes FJ. Chapter 37 - Aging and Reproduction. In: Plant TM, Zeleznik AJ, editors. Knobil and Neill's Physiology of Reproduction (Fourth Edition). San Diego: Academic Press; 2015. p. 1661-93. 505. McKenna NJ. Chapter 9 - Gonadal Steroid Action. In: Plant TM, Zeleznik AJ, editors. Knobil and Neill's Physiology of Reproduction (Fourth Edition). San Diego: Academic Press; 2015. p. 313-33. 506. McArdle CA, Roberson MS. Chapter 10 - Gonadotropes and Gonadotropin- Releasing Hormone Signaling. In: Plant TM, Zeleznik AJ, editors. Knobil and Neill's Physiology of Reproduction (Fourth Edition). San Diego: Academic Press; 2015. p. 335- 97. 507. Pangas SA, Rajkovic A. Chapter 21 - Follicular Development: Mouse, Sheep, and Human Models. In: Plant TM, Zeleznik AJ, editors. Knobil and Neill's Physiology of Reproduction (Fourth Edition). San Diego: Academic Press; 2015. p. 947-95. 508. Gallicchio L, Flaws JA, Smith RL. Age at menarche, androgen concentrations, and midlife obesity: findings from the Midlife Women's Health Study. Menopause. 2016;23(11):1182-8. Epub 2016/10/26. doi: 10.1097/gme.0000000000000691. PubMed PMID: 27433862; PMCID: PMC5079777. 509. Miller SR, Gallicchio LM, Lewis LM, Babus JK, Langenberg P, Zacur HA, Flaws JA. Association between race and hot flashes in midlife women. Maturitas. 2006;54(3):260-9. Epub 2006/01/21. doi: 10.1016/j.maturitas.2005.12.001. PubMed PMID: 16423474. 510. Smith RL, Gallicchio LM, Flaws JA. Understanding the complex relationships underlying hot flashes: a Bayesian network approach. Menopause. 2018;25(2):182-90. doi: 10.1097/GME.0000000000000959. PubMed PMID: 28763402; PMCID: PMC5771829. 511. Gallicchio L, Schilling C, Miller SR, Zacur H, Flaws JA. Correlates of depressive symptoms among women undergoing the menopausal transition. J Psychosom Res. 2007;63(3):263-8. Epub 2007/08/28. doi: 10.1016/j.jpsychores.2007.02.003. PubMed PMID: 17719363. 512. Bao A-M, Liu R-Y, van Someren EJW, Hofman MA, Cao Y-X, Zhou J-N. Diurnal rhythm of free estradiol during the menstrual cycle. European journal of endocrinology. 2003;148(2):227-32. doi: 10.1530/eje.0.1480227. PubMed PMID: 12590642. 201 513. Vermeulen A. The hormonal activity of the postmenopausal ovary. J Clin Endocrinol Metab. 1976;42(2):247-53. doi: 10.1210/jcem-42-2-247. PubMed PMID: 177438. 514. Reinsberg J, Wegener-Toper P, van der Ven K, van der Ven H, Klingmueller D. Effect of mono-(2-ethylhexyl) phthalate on steroid production of human granulosa cells. Toxicol Appl Pharmacol. 2009;239(1):116-23. Epub 2009/06/09. doi: 10.1016/j.taap.2009.05.022. PubMed PMID: 19501113. 515. Araki A, Mitsui T, Miyashita C, Nakajima T, Naito H, Ito S, Sasaki S, Cho K, Ikeno T, Nonomura K, Kishi R. Association between maternal exposure to di(2- ethylhexyl) phthalate and reproductive hormone levels in fetal blood: the Hokkaido study on environment and children's health. PLoS One. 2014;9(10):e109039. doi: 10.1371/journal.pone.0109039. PubMed PMID: 25296284; PMCID: PMC4189794. 516. Sathyanarayana S, Butts S, Wang C, Barrett E, Nguyen R, Schwartz SM, Haaland W, Swan SH, Team T. Early Prenatal Phthalate Exposure, Sex Steroid Hormones, and Birth Outcomes. J Clin Endocrinol Metab. 2017;102(6):1870-8. doi: 10.1210/jc.2016-3837. PubMed PMID: 28324030. 517. Cao M, Pan W, Shen X, Li C, Zhou J, Liu J. Urinary levels of phthalate metabolites in women associated with risk of premature ovarian failure and reproductive hormones. Chemosphere. 2020;242:125206. Epub 2019/11/05. doi: 10.1016/j.chemosphere.2019.125206. PubMed PMID: 31678849. 518. Hannon PR, Brannick KE, Wang W, Gupta RK, Flaws JA. Di(2-ethylhexyl) phthalate inhibits antral follicle growth, induces atresia, and inhibits steroid hormone production in cultured mouse antral follicles. Toxicol Appl Pharmacol. 2015;284(1):42- 53. doi: 10.1016/j.taap.2015.02.010. PubMed PMID: 25701202; PMCID: PMC4374011. 519. Du Y, Guo N, Wang Y, Teng X, Hua X, Deng T, Yao Y, Yuan X, Li Y. Follicular fluid concentrations of phthalate metabolites are associated with altered intrafollicular reproductive hormones in women undergoing in vitro fertilization. Fertility and sterility. 2019;111(5):953-61. Epub 2019/03/19. doi: 10.1016/j.fertnstert.2019.01.021. PubMed PMID: 30879714. 520. Du Y-Y, Guo N, Wang Y-X, Hua X, Deng T-R, Teng X-M, Yao Y-C, Li Y-F. Urinary phthalate metabolites in relation to serum anti-Müllerian hormone and inhibin B levels among women from a fertility center: a retrospective analysis. Reprod Health. 2018;15(1):33-. doi: 10.1186/s12978-018-0469-8. PubMed PMID: 29471860. 521. Zhang J, Yin W, Li P, Hu C, Wang L, Li T, Gao E, Hou J, Wang G, Wang X, Wang L, Yu Z, Yuan J. Interaction between diet- and exercise-lifestyle and phthalates exposure on sex hormone levels. J Hazard Mater. 2019;369:290-8. doi: 10.1016/j.jhazmat.2019.02.011. PubMed PMID: 30780025. 522. Watkins DJ, Sanchez BN, Tellez-Rojo MM, Lee JM, Mercado-Garcia A, Blank- Goldenberg C, Peterson KE, Meeker JD. Phthalate and bisphenol A exposure during in 202 utero windows of susceptibility in relation to reproductive hormones and pubertal development in girls. Environ Res. 2017;159:143-51. doi: 10.1016/j.envres.2017.07.051. PubMed PMID: 28800472; PMCID: PMC5623649. 523. Johns LE, Ferguson KK, Soldin OP, Cantonwine DE, Rivera-González LO, Del Toro LVA, Calafat AM, Ye X, Alshawabkeh AN, Cordero JF, Meeker JD. Urinary phthalate metabolites in relation to maternal serum thyroid and sex hormone levels during pregnancy: a longitudinal analysis. Reproductive biology and endocrinology : RB&E. 2015;13:4-. doi: 10.1186/1477-7827-13-4. PubMed PMID: 25596636. 524. Hart R, Doherty DA, Frederiksen H, Keelan JA, Hickey M, Sloboda D, Pennell CE, Newnham JP, Skakkebaek NE, Main KM. The influence of antenatal exposure to phthalates on subsequent female reproductive development in adolescence: a pilot study. Reproduction. 2013. Epub 2013/09/13. doi: 10.1530/rep-13-0331. PubMed PMID: 24025997. 525. Li N, Liu T, Guo K, Zhu J, Yu G, Wang S, Ye L. Effect of mono-(2-ethylhexyl) phthalate (MEHP) on proliferation of and steroid hormone synthesis in rat ovarian granulosa cells in vitro. J Cell Physiol. 2018;233(4):3629-37. Epub 2017/10/17. doi: 10.1002/jcp.26224. PubMed PMID: 29034469. 526. Meling DD, Warner GR, Szumski JR, Gao L, Gonsioroski AV, Rattan S, Flaws JA. The effects of a phthalate metabolite mixture on antral follicle growth and sex steroid synthesis in mice. Toxicol Appl Pharmacol. 2020;388:114875. Epub 2019/12/31. doi: 10.1016/j.taap.2019.114875. PubMed PMID: 31884101; PMCID: PMC7017935. 527. Ma Y, Zhang J, Zeng R, Qiao X, Cheng R, Nie Y, Luo Y, Li S, Zhang J, Xu W, Xu L, Hu Y. Effects of the Dibutyl Phthalate (DBP) on the Expression and Activity of Aromatase in Human Granulosa Cell Line KGN. Ann Clin Lab Sci. 2019;49(2):175-82. Epub 2019/04/28. PubMed PMID: 31028061. 528. Kershaw EE, Flier JS. Adipose Tissue as an Endocrine Organ. The Journal of Clinical Endocrinology & Metabolism. 2004;89(6):2548-56. doi: 10.1210/jc.2004-0395. 529. Kim SH, Park MJ. Phthalate exposure and childhood obesity. Ann Pediatr Endocrinol Metab. 2014;19(2):69-75. Epub 2014/08/01. doi: 10.6065/apem.2014.19.2.69. PubMed PMID: 25077088; PMCID: PMC4114051. 530. Zettergren A, Andersson N, Larsson K, Kull I, Melén E, Georgelis A, Berglund M, Lindh C, Bergström A. Exposure to environmental phthalates during preschool age and obesity from childhood to young adulthood. Environ Res. 2021;192:110249. Epub 2020/09/28. doi: 10.1016/j.envres.2020.110249. PubMed PMID: 32980305. 531. Manson JM, Sammel MD, Freeman EW, Grisso JA. Racial differences in sex hormone levels in women approaching the transition to menopause. Fertil Steril. 2001;75(2):297-304. Epub 2001/02/15. doi: 10.1016/s0015-0282(00)01723-4. PubMed PMID: 11172830. 203 532. Pinheiro SP, Holmes MD, Pollak MN, Barbieri RL, Hankinson SE. Racial differences in premenopausal endogenous hormones. Cancer Epidemiol Biomarkers Prev. 2005;14(9):2147-53. Epub 2005/09/21. doi: 10.1158/1055-9965.Epi-04-0944. PubMed PMID: 16172224. 533. Setiawan VW, Haiman CA, Stanczyk FZ, Le Marchand L, Henderson BE. Racial/ethnic differences in postmenopausal endogenous hormones: the multiethnic cohort study. Cancer Epidemiol Biomarkers Prev. 2006;15(10):1849-55. Epub 2006/10/13. doi: 10.1158/1055-9965.Epi-06-0307. PubMed PMID: 17035391. 534. Schliep KC, Mumford SL, Vladutiu CJ, Ahrens KA, Perkins NJ, Sjaarda LA, Kissell KA, Prasad A, Wactawski-Wende J, Schisterman EF. Perceived stress, reproductive hormones, and ovulatory function: a prospective cohort study. Epidemiology. 2015;26(2):177-84. Epub 2015/02/03. doi: 10.1097/ede.0000000000000238. PubMed PMID: 25643098; PMCID: PMC4315337. 535. Serrano SE, Braun J, Trasande L, Dills R, Sathyanarayana S. Phthalates and diet: a review of the food monitoring and epidemiology data. Environmental Health. 2014;13(1):43. doi: 10.1186/1476-069X-13-43. 536. Varshavsky JR, Morello-Frosch R, Woodruff TJ, Zota AR. Dietary sources of cumulative phthalates exposure among the U.S. general population in NHANES 2005- 2014. Environment international. 2018;115:417-29. Epub 2018/03/29. doi: 10.1016/j.envint.2018.02.029. PubMed PMID: 29605141. 537. Carruba G, Granata OM, Pala V, Campisi I, Agostara B, Cusimano R, Ravazzolo B, Traina A. A traditional Mediterranean diet decreases endogenous estrogens in healthy postmenopausal women. Nutr Cancer. 2006;56(2):253-9. Epub 2007/05/04. doi: 10.1207/s15327914nc5602_18. PubMed PMID: 17474873. 538. Chang VC, Cotterchio M, Boucher BA, Jenkins DJA, Mirea L, McCann SE, Thompson LU. Effect of Dietary Flaxseed Intake on Circulating Sex Hormone Levels among Postmenopausal Women: A Randomized Controlled Intervention Trial. Nutr Cancer. 2019;71(3):385-98. Epub 2018/10/31. doi: 10.1080/01635581.2018.1516789. PubMed PMID: 30375890. 539. Freedman RR. Physiology of hot flashes. Am J Hum Biol. 2001;13(4):453-64. doi: 10.1002/ajhb.1077. PubMed PMID: 11400216. 540. Kronenberg F. Hot flashes: epidemiology and physiology. Ann N Y Acad Sci. 1990;592:52-86; discussion 123-33. doi: 10.1111/j.1749-6632.1990.tb30316.x. PubMed PMID: 2197954. 541. Sarrel P, Portman D, Lefebvre P, Lafeuille MH, Grittner AM, Fortier J, Gravel J, Duh MS, Aupperle PM. Incremental direct and indirect costs of untreated vasomotor symptoms. Menopause. 2015;22(3):260-6. Epub 2015/02/26. doi: 10.1097/GME.0000000000000320. PubMed PMID: 25714236. 204 542. Ziv-Gal A, Flaws JA. Factors that may influence the experience of hot flushes by healthy middle-aged women. J Womens Health (Larchmt). 2010;19(10):1905-14. doi: 10.1089/jwh.2009.1852. PubMed PMID: 20831431; PMCID: PMC2965699. 543. Ziv-Gal A, Gallicchio L, Chiang C, Ther SN, Miller SR, Zacur HA, Dills RL, Flaws JA. Phthalate metabolite levels and menopausal hot flashes in midlife women. Reproductive Toxicology. 2016;60:76-81. doi: https://doi.org/10.1016/j.reprotox.2016.02.001. 544. Heudorf U, Mersch-Sundermann V, Angerer J. Phthalates: toxicology and exposure. Int J Hyg Environ Health. 2007;210(5):623-34. doi: 10.1016/j.ijheh.2007.07.011. PubMed PMID: 17889607. 545. Biesterbos JW, Dudzina T, Delmaar CJ, Bakker MI, Russel FG, von Goetz N, Scheepers PT, Roeleveld N. Usage patterns of personal care products: important factors for exposure assessment. Food Chem Toxicol. 2013;55:8-17. doi: 10.1016/j.fct.2012.11.014. PubMed PMID: 23174517. 546. Silva MJ, Barr DB, Reidy JA, Malek NA, Hodge CC, Caudill SP, Brock JW, Needham LL, Calafat AM. Urinary levels of seven phthalate metabolites in the U.S. population from the National Health and Nutrition Examination Survey (NHANES) 1999- 2000. Environ Health Perspect. 2004;112(3):331-8. doi: 10.1289/ehp.6723. PubMed PMID: 14998749; PMCID: PMC1241863. 547. Davis BJ, Weaver R, Gaines LJ, Heindel JJ. Mono-(2-ethylhexyl) phthalate suppresses estradiol production independent of FSH-cAMP stimulation in rat granulosa cells. Toxicol Appl Pharmacol. 1994;128(2):224-8. doi: 10.1006/taap.1994.1201. PubMed PMID: 7940537. 548. Hannon PR, Brannick KE, Wang W, Flaws JA. Mono(2-ethylhexyl) phthalate accelerates early folliculogenesis and inhibits steroidogenesis in cultured mouse whole ovaries and antral follicles. Biol Reprod. 2015;92(5):120. Epub 2015/03/27. doi: 10.1095/biolreprod.115.129148. PubMed PMID: 25810477; PMCID: PMC4645979. 549. Lake BG, Gray TJ, Lewis DF, Beamand JA, Hodder KD, Purchase R, Gangolli SD. Structure-activity relationships for induction of peroxisomal enzyme activities by phthalate monoesters in primary rat hepatocyte cultures. Toxicol Ind Health. 1987;3(2):165-83. doi: 10.1177/074823378700300212. PubMed PMID: 3617066. 550. Lovekamp-Swan T, Davis BJ. Mechanisms of phthalate ester toxicity in the female reproductive system. Environ Health Perspect. 2003;111(2):139-45. doi: 10.1289/ehp.5658. PubMed PMID: 12573895; PMCID: PMC1241340. 551. Lovekamp TN, Davis BJ. Mono-(2-ethylhexyl) phthalate suppresses aromatase transcript levels and estradiol production in cultured rat granulosa cells. Toxicol Appl Pharmacol. 2001;172(3):217-24. doi: 10.1006/taap.2001.9156. PubMed PMID: 11312650. 205 552. Wang W, Craig ZR, Basavarajappa MS, Hafner KS, Flaws JA. Mono-(2- ethylhexyl) phthalate induces oxidative stress and inhibits growth of mouse ovarian antral follicles. Biol Reprod. 2012;87(6):152. doi: 10.1095/biolreprod.112.102467. PubMed PMID: 23077170; PMCID: PMC4435432. 553. Wang YX, Zeng Q, Sun Y, You L, Wang P, Li M, Yang P, Li J, Huang Z, Wang C, Li S, Dan Y, Li YF, Lu WQ. Phthalate exposure in association with serum hormone levels, sperm DNA damage and spermatozoa apoptosis: A cross-sectional study in China. Environ Res. 2016;150:557-65. doi: 10.1016/j.envres.2015.11.023. PubMed PMID: 26654563. 554. Meeker JD, Missmer SA, Altshul L, Vitonis AF, Ryan L, Cramer DW, Hauser R. Serum and follicular fluid organochlorine concentrations among women undergoing assisted reproduction technologies. Environ Health. 2009;8:32. doi: 10.1186/1476- 069X-8-32. PubMed PMID: 19594949; PMCID: PMC2717935. 555. Zhou C, Flaws JA. Effects of an Environmentally Relevant Phthalate Mixture on Cultured Mouse Antral Follicles. Toxicol Sci. 2017;156(1):217-29. doi: 10.1093/toxsci/kfw245. PubMed PMID: 28013214; PMCID: PMC6075604. 556. Sturdee DW. The menopausal hot flush--anything new? Maturitas. 2008;60(1):42-9. doi: 10.1016/j.maturitas.2008.02.006. PubMed PMID: 18384981. 557. Gallicchio L, Miller SR, Kiefer J, Greene T, Zacur HA, Flaws JA. Risk factors for hot flashes among women undergoing the menopausal transition: baseline results from the Midlife Women's Health Study. Menopause. 2015;22(10):1098-107. Epub 2015/03/19. doi: 10.1097/GME.0000000000000434. PubMed PMID: 25783472; PMCID: PMC4573383. 558. Harlow SD, Gass M, Hall JE, Lobo R, Maki P, Rebar RW, Sherman S, Sluss PM, de Villiers TJ, Group SC. Executive summary of the Stages of Reproductive Aging Workshop +10: addressing the unfinished agenda of staging reproductive aging. Climacteric. 2012;15(2):105-14. doi: 10.3109/13697137.2011.650656. PubMed PMID: 22338612; PMCID: PMC3580996. 559. Radloff LS. The CES-D Scale:A Self-Report Depression Scale for Research in the General Population. Applied Psychological Measurement. 1977;1(3):385-401. doi: 10.1177/014662167700100306. 560. Cochran CJ, Gallicchio L, Miller SR, Zacur H, Flaws JA. Cigarette smoking, androgen levels, and hot flushes in midlife women. Obstet Gynecol. 2008;112(5):1037- 44. Epub 2008/11/04. doi: 10.1097/AOG.0b013e318189a8e2. PubMed PMID: 18978103; PMCID: PMC2673540. 561. Gallicchio L, Miller SR, Visvanathan K, Lewis LM, Babus J, Zacur H, Flaws JA. Cigarette smoking, estrogen levels, and hot flashes in midlife women. Maturitas. 2006;53(2):133-43. doi: 10.1016/j.maturitas.2005.03.007. PubMed PMID: 16368467. 206 562. Visvanathan K, Gallicchio L, Schilling C, Babus JK, Lewis LM, Miller SR, Zacur H, Flaws JA. Cytochrome gene polymorphisms, serum estrogens, and hot flushes in midlife women. Obstet Gynecol. 2005;106(6):1372-81. doi: 10.1097/01.AOG.0000187308.67021.98. PubMed PMID: 16319265. 563. Ziv-Gal A, Gallicchio L, Miller SR, Zacur HA, Flaws JA. Genetic polymorphisms in the aryl hydrocarbon receptor signaling pathway as potential risk factors of menopausal hot flashes. Am J Obstet Gynecol. 2012;207(3):202 e9- e18. doi: 10.1016/j.ajog.2012.05.019. PubMed PMID: 22840970; PMCID: PMC3432652. 564. Hogberg J, Hanberg A, Berglund M, Skerfving S, Remberger M, Calafat AM, Filipsson AF, Jansson B, Johansson N, Appelgren M, Hakansson H. Phthalate diesters and their metabolites in human breast milk, blood or serum, and urine as biomarkers of exposure in vulnerable populations. Environ Health Perspect. 2008;116(3):334-9. doi: 10.1289/ehp.10788. PubMed PMID: 18335100; PMCID: PMC2265037. 565. Staples CA, Peterson DR, Parkerton TF, Adams WJ. The environmental fate of phthalate esters: A literature review. Chemosphere. 1997;35(4):667-749. doi: https://doi.org/10.1016/S0045-6535(97)00195-1. 566. Wolff MS, Engel SM, Berkowitz GS, Ye X, Silva MJ, Zhu C, Wetmur J, Calafat AM. Prenatal phenol and phthalate exposures and birth outcomes. Environ Health Perspect. 2008;116(8):1092-7. doi: 10.1289/ehp.11007. PubMed PMID: 18709157; PMCID: PMC2516577. 567. Radke EG, Braun JM, Meeker JD, Cooper GS. Phthalate exposure and male reproductive outcomes: A systematic review of the human epidemiological evidence. Environ Int. 2018;121(Pt 1):764-93. doi: 10.1016/j.envint.2018.07.029. PubMed PMID: 30336412. 568. Radke EG, Glenn BS, Braun JM, Cooper GS. Phthalate exposure and female reproductive and developmental outcomes: a systematic review of the human epidemiological evidence. Environ Int. 2019;130:104580. doi: 10.1016/j.envint.2019.02.003. PubMed PMID: 31351310; PMCID: PMC9400136. 569. Ferguson KK, Rosen EM, Rosario Z, Feric Z, Calafat AM, McElrath TF, Velez Vega C, Cordero JF, Alshawabkeh A, Meeker JD. Environmental phthalate exposure and preterm birth in the PROTECT birth cohort. Environ Int. 2019;132:105099. doi: 10.1016/j.envint.2019.105099. PubMed PMID: 31430608; PMCID: PMC6754790. 570. Johns LE, Cooper GS, Galizia A, Meeker JD. Exposure assessment issues in epidemiology studies of phthalates. Environ Int. 2015;85:27-39. doi: 10.1016/j.envint.2015.08.005. PubMed PMID: 26313703; PMCID: PMC4648682. 571. CDC. Metabolites of phthalates and phthalate alternatives: Urine 2013. Available from: https://www.cdc.gov/nchs/data/nhanes/nhanes_11_12/PHTHTE_G_met.pdf. 207 572. Phthalates and Hot Flashes SI [Internet]. University of Illinois at Urbana- Champaign. 2020 [cited 10/20/2021]. Available from: /10.13012/B2IDB-9238850_V1. 573. Marie C, Vendittelli F, Sauvant-Rochat MP. Obstetrical outcomes and biomarkers to assess exposure to phthalates: A review. Environ Int. 2015;83:116-36. doi: 10.1016/j.envint.2015.06.003. PubMed PMID: 26118330. 574. Gallicchio L, Visvanathan K, Miller SR, Babus J, Lewis LM, Zacur H, Flaws JA. Body mass, estrogen levels, and hot flashes in midlife women. Am J Obstet Gynecol. 2005;193(4):1353-60. Epub 2005/10/06. doi: 10.1016/j.ajog.2005.04.001. PubMed PMID: 16202725. 575. Edwards BJ, Li J. Endocrinology of menopause. Periodontol 2000. 2013;61(1):177-94. doi: 10.1111/j.1600-0757.2011.00407.x. PubMed PMID: 23240949. 576. Chiang C, Lewis LR, Borkowski G, Flaws JA. Exposure to di(2-ethylhexyl) phthalate and diisononyl phthalate during adulthood disrupts hormones and ovarian folliculogenesis throughout the prime reproductive life of the mouse. Toxicol Appl Pharmacol. 2020;393:114952. doi: 10.1016/j.taap.2020.114952. PubMed PMID: 32165126; PMCID: PMC7138141. 577. Hannon PR, Niermann S, Flaws JA. Acute Exposure to Di(2-Ethylhexyl) Phthalate in Adulthood Causes Adverse Reproductive Outcomes Later in Life and Accelerates Reproductive Aging in Female Mice. Toxicol Sci. 2016;150(1):97-108. doi: 10.1093/toxsci/kfv317. PubMed PMID: 26678702; PMCID: PMC5009616. 578. Grindler NM, Allsworth JE, Macones GA, Kannan K, Roehl KA, Cooper AR. Persistent organic pollutants and early menopause in U.S. women. PLoS One. 2015;10(1):e0116057. doi: 10.1371/journal.pone.0116057. PubMed PMID: 25629726; PMCID: PMC4309567. 579. Ozel S, Tokmak A, Aykut O, Aktulay A, Hancerliogullari N, Engin Ustun Y. Serum levels of phthalates and bisphenol-A in patients with primary ovarian insufficiency. Gynecol Endocrinol. 2019;35(4):364-7. Epub 2019/01/15. doi: 10.1080/09513590.2018.1534951. PubMed PMID: 30638094. 580. Gibson-Helm M, Teede H, Vincent A. Symptoms, health behavior and understanding of menopause therapy in women with premature menopause. Climacteric. 2014;17(6):666-73. doi: 10.3109/13697137.2014.913284. PubMed PMID: 24742007. 581. Du YY, Guo N, Wang YX, Hua X, Deng TR, Teng XM, Yao YC, Li YF. Urinary phthalate metabolites in relation to serum anti-Mullerian hormone and inhibin B levels among women from a fertility center: a retrospective analysis. Reprod Health. 2018;15(1):33. doi: 10.1186/s12978-018-0469-8. PubMed PMID: 29471860; PMCID: PMC5824533. 208 582. Bansal R, Aggarwal N. Menopausal Hot Flashes: A Concise Review. J Midlife Health. 2019;10(1):6-13. doi: 10.4103/jmh.JMH_7_19. PubMed PMID: 31001050; PMCID: PMC6459071. 583. Nelson HD. Commonly used types of postmenopausal estrogen for treatment of hot flashes: scientific review. JAMA. 2004;291(13):1610-20. doi: 10.1001/jama.291.13.1610. PubMed PMID: 15069049. 584. Smith RL, Gallicchio L, Miller SR, Zacur HA, Flaws JA. Risk Factors for Extended Duration and Timing of Peak Severity of Hot Flashes. PLoS One. 2016;11(5):e0155079. doi: 10.1371/journal.pone.0155079. PubMed PMID: 27149066; PMCID: PMC4858155. 585. Gupta RK, Singh JM, Leslie TC, Meachum S, Flaws JA, Yao HH. Di-(2- ethylhexyl) phthalate and mono-(2-ethylhexyl) phthalate inhibit growth and reduce estradiol levels of antral follicles in vitro. Toxicol Appl Pharmacol. 2010;242(2):224-30. doi: 10.1016/j.taap.2009.10.011. PubMed PMID: 19874833; PMCID: PMC2789888. 586. Kim MJ, Moon S, Oh BC, Jung D, Choi K, Park YJ. Association Between Diethylhexyl Phthalate Exposure and Thyroid Function: A Meta-Analysis. Thyroid. 2019;29(2):183-92. doi: 10.1089/thy.2018.0051. PubMed PMID: 30588877; PMCID: PMC6488044. 587. Meeker JD, Ferguson KK. Relationship between urinary phthalate and bisphenol A concentrations and serum thyroid measures in U.S. adults and adolescents from the National Health and Nutrition Examination Survey (NHANES) 2007-2008. Environ Health Perspect. 2011;119(10):1396-402. doi: 10.1289/ehp.1103582. PubMed PMID: 21749963; PMCID: PMC3230451. 588. James-Todd TM, Chiu YH, Zota AR. Racial/ethnic disparities in environmental endocrine disrupting chemicals and women's reproductive health outcomes: epidemiological examples across the life course. Current epidemiology reports. 2016;3(2):161-80. Epub 2017/05/13. doi: 10.1007/s40471-016-0073-9. PubMed PMID: 28497013; PMCID: PMC5423735. 589. Varshavsky JR, Zota AR, Woodruff TJ. A Novel Method for Calculating Potency- Weighted Cumulative Phthalates Exposure with Implications for Identifying Racial/Ethnic Disparities among U.S. Reproductive-Aged Women in NHANES 2001- 2012. Environ Sci Technol. 2016;50(19):10616-24. doi: 10.1021/acs.est.6b00522. PubMed PMID: 27579903; PMCID: PMC5748889. 590. Helm JS, Nishioka M, Brody JG, Rudel RA, Dodson RE. Measurement of endocrine disrupting and asthma-associated chemicals in hair products used by Black women. Environ Res. 2018;165:448-58. Epub 2018/05/01. doi: 10.1016/j.envres.2018.03.030. PubMed PMID: 29705122. 591. Natari RB, Clavarino AM, McGuire TM, Dingle KD, Hollingworth SA. The bidirectional relationship between vasomotor symptoms and depression across the 209 menopausal transition: a systematic review of longitudinal studies. Menopause. 2018;25(1):109-20. doi: 10.1097/GME.0000000000000949. PubMed PMID: 28719420. 592. Shiue I. Urinary heavy metals, phthalates and polyaromatic hydrocarbons independent of health events are associated with adult depression: USA NHANES, 2011-2012. Environ Sci Pollut Res Int. 2015;22(21):17095-103. doi: 10.1007/s11356- 015-4944-2. PubMed PMID: 26126689. 593. Kim KN, Choi YH, Lim YH, Hong YC. Urinary phthalate metabolites and depression in an elderly population: National Health and Nutrition Examination Survey 2005-2012. Environ Res. 2016;145:61-7. doi: 10.1016/j.envres.2015.11.021. PubMed PMID: 26624239. 594. Lee KS, Lim YH, Kim KN, Choi YH, Hong YC, Lee N. Urinary phthalate metabolites concentrations and symptoms of depression in an elderly population. Sci Total Environ. 2018;625:1191-7. doi: 10.1016/j.scitotenv.2017.12.219. PubMed PMID: 29996415. 595. Silva MJ, Samandar E, Reidy JA, Hauser R, Needham LL, Calafat AM. Metabolite profiles of di-n-butyl phthalate in humans and rats. Environ Sci Technol. 2007;41(21):7576-80. Epub 2007/11/30. doi: 10.1021/es071142x. PubMed PMID: 18044544. 596. Miller HG, Li RM. Measuring hot flashes: summary of a National Institutes of Health workshop. Mayo Clin Proc. 2004;79(6):777-81. doi: 10.4065/79.6.777. PubMed PMID: 15182093. 597. Al-Hendy A, Myers ER, Stewart E. Uterine Fibroids: Burden and Unmet Medical Need. Semin Reprod Med. 2017;35(6):473-80. doi: 10.1055/s-0037-1607264. PubMed PMID: 29100234; PMCID: PMC6193285. 598. Yu O, Scholes D, Schulze-Rath R, Grafton J, Hansen K, Reed SD. A US population-based study of uterine fibroid diagnosis incidence, trends, and prevalence: 2005 through 2014. Am J Obstet Gynecol. 2018;219(6):591 e1- e8. doi: 10.1016/j.ajog.2018.09.039. PubMed PMID: 30291840. 599. Marsh EE, Al-Hendy A, Kappus D, Galitsky A, Stewart EA, Kerolous M. Burden, Prevalence, and Treatment of Uterine Fibroids: A Survey of U.S. Women. J Womens Health (Larchmt). 2018;27(11):1359-67. doi: 10.1089/jwh.2018.7076. PubMed PMID: 30230950; PMCID: PMC6247381. 600. Wise LA, Laughlin-Tommaso SK. Epidemiology of Uterine Fibroids: From Menarche to Menopause. Clin Obstet Gynecol. 2016;59(1):2-24. doi: 10.1097/GRF.0000000000000164. PubMed PMID: 26744813; PMCID: PMC4733579. 601. Ulin M, Ali M, Chaudhry ZT, Al-Hendy A, Yang Q. Uterine fibroids in menopause and perimenopause. Menopause. 2020;27(2):238-42. doi: 10.1097/GME.0000000000001438. PubMed PMID: 31834160; PMCID: PMC6994343. 210 602. De La Cruz MS, Buchanan EM. Uterine Fibroids: Diagnosis and Treatment. Am Fam Physician. 2017;95(2):100-7. PubMed PMID: 28084714. 603. Giuliani E, As-Sanie S, Marsh EE. Epidemiology and management of uterine fibroids. Int J Gynaecol Obstet. 2020;149(1):3-9. doi: 10.1002/ijgo.13102. PubMed PMID: 31960950. 604. Prevention CfDCa. Fourth national report on human exposure to environmental chemicals, updated tables. Atlanta, Georgia: US Department of Health and Human Services Centers for Disease Control and Prevention, 2019. 605. Sirohi D, Al Ramadhani R, Knibbs LD. Environmental exposures to endocrine disrupting chemicals (EDCs) and their role in endometriosis: a systematic literature review. Rev Environ Health. 2021;36(1):101-15. doi: 10.1515/reveh-2020-0046. PubMed PMID: 32903210. 606. James-Todd TM, Huang T, Seely EW, Saxena AR. The association between phthalates and metabolic syndrome: the National Health and Nutrition Examination Survey 2001-2010. Environ Health. 2016;15:52. Epub 2016/04/16. doi: 10.1186/s12940- 016-0136-x. PubMed PMID: 27079661; PMCID: PMC4832560. 607. Rocha PRS, Oliveira VD, Vasques CI, Dos Reis PED, Amato AA. Exposure to endocrine disruptors and risk of breast cancer: A systematic review. Crit Rev Oncol Hematol. 2021;161:103330. doi: 10.1016/j.critrevonc.2021.103330. PubMed PMID: 33862246. 608. Fu Z, Zhao F, Chen K, Xu J, Li P, Xia D, Wu Y. Association between urinary phthalate metabolites and risk of breast cancer and uterine leiomyoma. Reprod Toxicol. 2017;74:134-42. doi: 10.1016/j.reprotox.2017.09.009. PubMed PMID: 28951174. 609. Maloney EK, Waxman DJ. trans-Activation of PPARalpha and PPARgamma by structurally diverse environmental chemicals. Toxicol Appl Pharmacol. 1999;161(2):209- 18. doi: 10.1006/taap.1999.8809. PubMed PMID: 10581215. 610. Gray LE, Ostby J, Furr J, Wolf CJ, Lambright C, Parks L, Veeramachaneni DN, Wilson V, Price M, Hotchkiss A, Orlando E, Guillette L. Effects of environmental antiandrogens on reproductive development in experimental animals. Hum Reprod Update. 2001;7(3):248-64. doi: 10.1093/humupd/7.3.248. PubMed PMID: 11392371. 611. Zota AR, Geller RJ, VanNoy BN, Marfori CQ, Tabbara S, Hu LY, Baccarelli AA, Moawad GN. Phthalate Exposures and MicroRNA Expression in Uterine Fibroids: The FORGE Study. Epigenet Insights. 2020;13:2516865720904057. doi: 10.1177/2516865720904057. PubMed PMID: 32128507; PMCID: PMC7031793. 612. Kim JH, Kim SH, Oh YS, Ihm HJ, Chae HD, Kim CH, Kang BM. In vitro effects of phthalate esters in human myometrial and leiomyoma cells and increased urinary level of phthalate metabolite in women with uterine leiomyoma. Fertil Steril. 211 2017;107(4):1061-9 e1. Epub 2017/03/16. doi: 10.1016/j.fertnstert.2017.01.015. PubMed PMID: 28292620. 613. Kim JH. Analysis of the in vitro effects of di-(2-ethylhexyl) phthalate exposure on human uterine leiomyoma cells. Exp Ther Med. 2018;15(6):4972-8. Epub 2018/05/29. doi: 10.3892/etm.2018.6040. PubMed PMID: 29805520; PMCID: PMC5958751. 614. Zota AR, Geller RJ, Calafat AM, Marfori CQ, Baccarelli AA, Moawad GN. Phthalates exposure and uterine fibroid burden among women undergoing surgical treatment for fibroids: a preliminary study. Fertil Steril. 2019;111(1):112-21. doi: 10.1016/j.fertnstert.2018.09.009. PubMed PMID: 30447935; PMCID: PMC6321778. 615. Lee JE, Song S, Cho E, Jang HJ, Jung H, Lee HY, Kim S, Kim O, Lee JE. Weight change and risk of uterine leiomyomas: Korea Nurses' Health Study. Curr Med Res Opin. 2018;34(11):1913-9. doi: 10.1080/03007995.2018.1462783. PubMed PMID: 29625536. 616. Marie C, Vendittelli F, Sauvant-Rochat MP. Obstetrical outcomes and biomarkers to assess exposure to phthalates: A review. Environment International. 2015;83:116-36. doi: 10.1016/j.envint.2015.06.003. PubMed PMID: 26118330. 617. Barros AJ, Hirakata VN. Alternatives for logistic regression in cross-sectional studies: an empirical comparison of models that directly estimate the prevalence ratio. BMC Med Res Methodol. 2003;3:21. doi: 10.1186/1471-2288-3-21. PubMed PMID: 14567763; PMCID: PMC521200. 618. About Adult BMI 2017 [cited 2019 March 13]. Available from: https://www.cdc.gov/healthyweight/assessing/bmi/adult_bmi/index.html. 619. Dewalque L, Pirard C, Vandepaer S, Charlier C. Temporal variability of urinary concentrations of phthalate metabolites, parabens and benzophenone-3 in a Belgian adult population. Environ Res. 2015;142:414-23. doi: 10.1016/j.envres.2015.07.015. PubMed PMID: 26233661. 620. Qin H, Lin Z, Vasquez E, Luan X, Guo F, Xu L. Association between obesity and the risk of uterine fibroids: a systematic review and meta-analysis. J Epidemiol Community Health. 2021;75(2):197-204. doi: 10.1136/jech-2019-213364. PubMed PMID: 33067250. 621. Droyvold WB, Nilsen TI, Kruger O, Holmen TL, Krokstad S, Midthjell K, Holmen J. Change in height, weight and body mass index: Longitudinal data from the HUNT Study in Norway. Int J Obes (Lond). 2006;30(6):935-9. doi: 10.1038/sj.ijo.0803178. PubMed PMID: 16418765. 622. Myers SL, Baird DD, Olshan AF, Herring AH, Schroeder JC, Nylander-French LA, Hartmann KE. Self-report versus ultrasound measurement of uterine fibroid status. J Womens Health (Larchmt). 2012;21(3):285-93. Epub 2011/11/03. doi: 10.1089/jwh.2011.3008. PubMed PMID: 22044079; PMCID: PMC3298676. 212 623. Pavone D, Clemenza S, Sorbi F, Fambrini M, Petraglia F. Epidemiology and Risk Factors of Uterine Fibroids. Best Pract Res Clin Obstet Gynaecol. 2018;46:3-11. doi: 10.1016/j.bpobgyn.2017.09.004. PubMed PMID: 29054502. 213 APPENDIX A: DIETARY PREDICTORS OF PHTHALATE AND BISPHENOL EXPOSURES IN PREGNANT WOMEN This article/appendix has been published in Advances in Nutrition; Volume 10 Issue 5; Pacyga DC, Sathyanarayana S, Strakovsky RS; Dietary Predictors of Phthalate and Bisphenol Exposures in Pregnant Women. Copyright Elsevier (2019); https://doi.org/10.1093/advances/nmz029. A.1. ABSTRACT Endocrine disrupting chemicals (EDCs) can disrupt fetal developmental processes during pregnancy, leading to long-term adverse outcomes in humans. A major source of exposure to EDCs, such as phthalates and bisphenols, is the food supply, primarily due to contamination from processing and packaging. Therefore, this review aimed to 1) review food-monitoring sources of phthalates and bisphenols, and 2) evaluate methodologies and provide future directions needed to establish EDC-limiting dietary recommendations in pregnancy. Using PubMed, 10 peer-reviewed studies were found on dietary predictors of EDC exposure in pregnancy, and all were selected for review. Use of plastic containers in pregnancy was associated with higher urinary phthalate metabolites, whereas canned food consumption was associated with higher urinary bisphenol A (BPA) concentrations. Foods and dietary patterns associated with healthier food choices (e.g., organic/grown/raised/caught foods, folic acid supplements, vegetarianism) were generally associated with lower urinary phthalate metabolite and BPA concentrations. Despite the many food-monitoring studies reporting high BPA and phthalate concentrations in various foods, the designs of most studies described here 214 were not sufficiently robust to consistently detect associations of specific foods/food groups with phthalates and BPA. Given the limitations of currently available research, future studies should incorporate more valid questionnaires to accurately assess dietary EDC exposure, strive for concurrent diet and exposure assessment, and assess whether geographical and cultural differences modify associations of diet with gestational EDC exposures. Such progress will be critical for developing dietary recommendations that ensure the safety and health of pregnant women. A.2. KEYWORDS Pregnancy; phthalates; bisphenols; toxicology; diet. A.3. INTRODUCTION Many consumer products, including food contact materials, contain phthalates and bisphenols, resulting in widespread human exposure to these chemicals. Phthalates are diesters of phthalic acid (254, 255) that are classified into two categories based on their molecular weight: high molecular weight phthalates (HighMWPs) and low molecular weight phthalates (LowMWPs) (Table 17). HighMWPs are used as plasticizers in polyvinyl chloride products to make plastics flexible for building materials, medical devices, and food processing or packaging (153, 239, 256, 257), whereas LowMWPs are primarily used as solvents, fixatives, and adhesives in personal care products and cosmetics (153, 239, 256). Bisphenol A (BPA), and its replacements, bisphenol S (BPS) and bisphenol F (BPF), are used to manufacture polycarbonate plastics and epoxy resins for consumer and food product packaging, including canned foods (256, 258-261). 215 Due to their short half-lives (< 24 hours), exposures to phthalates and bisphenols are best characterized in urine (compared with blood) (262). Upon exposure, phthalates, specifically, are metabolized and excreted in urine, allowing for approximation of exposure by measuring urinary phthalate parent-specific metabolites (summarized in Table 17). For example, di(2-ethylhexyl) phthalate (DEHP) exposure is approximated by assessing the sum of its urinary metabolites (mono(2-ethylhexyl) phthalate, mono(2- ethyl-5-hydroxyhexyl) phthalate, mono(2-ethyl-5-carboxypentyl) phthalate, and mono(2- ethyl-5-oxohexyl) phthalate), whereas diethyl phthalate (DEP) exposure is approximated by measuring its major urinary metabolite [monoethyl phthalate (mEP)]. According to the 2013-2014 U.S. NHANES, most women of reproductive age (15-44 years of age) (263) have urinary concentrations of these chemicals that are above the laboratory levels of detection (phthalates: 88-100%, BPA: 96%, BPS: 88%, BPF: 66%) (264). This is concerning because phthalates and bisphenols are known endocrine disrupting chemicals (EDCs), associated with adverse health outcomes, especially in pregnancy (79). EDCs can alter, mimic, or disrupt the function of gestational hormones, such as thyroid hormone, estrogens, and androgens (75, 265, 266), making pregnancy especially sensitive to the actions of EDCs. Human epidemiological studies have shown that prenatal exposure to EDCs, specifically phthalates and bisphenols, is associated with adverse pregnancy (24, 28) and birth outcomes (29, 267), as well as childhood behavioral problems (46, 268), respiratory problems (49, 269), and obesity (270, 271). Diet is a ubiquitous source of chronic EDC exposure (272-274), because these chemicals have been shown to migrate from food contact materials (plastics, paper, metal, glass, and printing inks) that protect food from physical damage and microbial spoilage, thereby 216 affecting human health (275). Human exposure to EDCs from food can be attributed to various factors, including animal feeding practices, food production, processing, and packaging practices, as well as food storage conditions (276). Characterizing dietary sources of EDCs requires accurate assessment of both EDC concentrations and dietary intake history, and this aim is especially challenging during pregnancy. This is due to the many anatomical, physiological (e.g., increased renal function), and metabolic changes (277, 278) that occur in pregnancy, as well as the numerous pregnancy-related changes in dietary patterns, including diet quality and quantity (279, 280). Despite these challenges, characterizing dietary sources of these chemicals during pregnancy is important, as recommendations are needed to minimize exposure, while providing pregnant women with accessible and nutritious foods necessary to sustain a healthy pregnancy. To address these pregnancy-specific challenges, the aims of this review are to (1) briefly review food-monitoring sources of phthalates and bisphenols in the general population and (2) evaluate methodologies and provide future directions to help establish EDC-limiting dietary recommendations for pregnant women. 217 Table 17. Summary of reviewed phthalate parent compounds/metabolites, bisphenols, and their proposed sources Categorization Parent Compound (Name; Abbreviation) Metabolite (Name; Abbreviation) Exposure Sources High Molecular Di(2-ethylhexyl) phthalate; DEHP Mono(2-ethylhexyl) phthalate; mEHP • PVC plastics Weight Mono(2-ethyl-5-hydroxyhexyl) phthalate; mEHHP • Food packaging & processing Phthalate Mono(2-ethyl-5-oxohexyl) phthalate; mEOHP • Medical devices Mono(2-ethyl-5-carboxypentyl) phthalate; mECPP • Pharmaceutical coatings • Building materials Di-isononyl phthalate; DiNP Mono-isononyl phthalate; mNP/miNP • PVC plastics Monooxononyl phthalate; mONP • Food packaging Monocarboxyoctyl phthalate; mCOP • Building materials • Car interiors • Drinking straws Di-isodecyl phthalate; DiDP Monocarboxynonyl phthalate; mCNP • PVC plastics • Food packaging • Building materials • Car interiors • Swimming pools Di-n-octyl phthalate; DOP/DnOP Mono(3-carboxypropyl) phthalate; mCPP • PVC plastics • Food packaging • Building materials • Adhesives Benzylbutyl phthalate; BBzP Monobenzyl phthalate; mBzP • PVC plastics • Food packaging • Car care products • Some PCPs Low Molecular Diethyl phthalate; DEP Monoethyl phthalate; mEP • Fragrant PCPs: Weight perfumes/colognes, Phthalate deodorants, soaps, shampoos, lotions Di-n-butyl phthalate; DBP/DnBP Mono-n-butyl phthalate; mBP/mnBP • PCPs: nail polish, cosmetics Mono-hydroxybutyl phthalate; mHBP • Printing inks Di-iso-butyl phthalate; DiBP Mono-isobutyl phthalate; miBP • Pharmaceutical coatings Mono-hydroxyl-isobutyl phthalate; mHiBP • Insecticides Bisphenol Bisphenol A; BPA • Polycarbonate plastics and epoxy resins Bisphenol S; BPS • Food packaging: lining food cans, beverage containers Bisphenol F; BPF • Plastic dinnerware • Dental sealants • Thermal receipts References for phthalates (153, 239, 256), bisphenols (256, 258). PCPs: personal care products; PVC: polyvinyl chloride. 218 A.4. METHODS PubMed was searched using combinations of various keywords including: diet(ary) + pregnant or pregnancy + predictor(s) or variability or determinants or distribution, + endocrine disruptors or EDCs or phthalate(s) or bisphenol(s) or BPA. Studies were included if they assessed associations of consumption of foods or dietary patterns with EDC exposures in pregnancy. Based on our literature search, only 10 pregnancy cohort studies have evaluated dietary predictors of EDC exposures (summarized in Table 18 (90, 92, 93, 119, 137, 281-285)). Briefly, the 10 studies recruited ≥ 26 participants from 2003 to 2014; six cohorts were from the U.S./Puerto Rico, two from Spain, one from the Netherlands, and one from Australia. The following chemicals were assessed in these studies: HighMWPs [DEHP, di-isononyl phthalate (DiNP), di-isodecyl phthalate (DiDP), di-n-octyl phthalate (DOP), and benzylbutyl phthalate (BBzP)], LowMWPs [DEP, di-n- butyl phthalate (DBP), and di-iso-butyl phthalate (DiBP)], BPA, BPS, and BPF. One study investigated associations of foods with urinary paraben, benzophenone-3, triclosan, 2,4- dichlorophenol, and 2,5-dichlorophenol concentrations (119), but this review focuses on phthalates and bisphenols because diet is not a major source of exposure to these other chemicals (256, 258, 286, 287). The food categories in Tables 19 and 20 were selected after abstracting all foods or dietary patterns from the 10 studies, and collapsing them across categories that were common to several studies (when possible). Additional subcategories were created when packaging or processing information was available for the same food item. For example, several studies reported on fish intake, but these studies assessed either general seafood intake, canned fish intake, or fish intake (unspecified type). 219 Table 18. Summary of studies assessing dietary predictors of phthalate and bisphenol exposures in pregnant women Study Name Recruitment, Urinary Chemical Chemicals Assessed Urine Samples Chemical Analysis Foods Assessed Dietary Assessment (Reference) Location, N Adjustment & Covariates New Jersey • 2003-2004 Phthalates: 1 urine • HPLC-MS/MS (CDC) • Microwaved foods (plastic • Questionnaire (at delivery, • None Cohort (137) • New Jersey • mCPP, mBzP • Before delivery containers) about pregnancy) • N = 150 • mEP, mBP, miBP • Plastic tableware • Plastic container storage Generation R • 2004-2005 Phthalates: 1 urine • HPLC-ESI-MS/MS • Folic acid supplement • 3-month semi-quantitative • Urinary creatinine (as Study (93) • Netherlands • Sum-HighMWPs = • 1st trimester (The New York State • Vegetables FFQ (1st trimester) covariate) • N = 642 mECPP+mEHHP+mEOHP+mCMHP+ Department of Health) • Grains • Lifestyle questionnaire: • Maternal age mCPP+mBzP+mHxP+mHpP • Fish/shellfish supplements in pregnancy • Ethnicity • Sum-DEHP = mECPP+mEHHP+mEOHP+mCMHP • Soft drinks (1st trimester) • Pre-pregnancy BMI • DOP: mCPP • Soups & bouillon • Education • Sum-LowMWPs = mMP+mEP+mBP+miBP • Daily dietary caloric intake • Parity • Phthalic acid • Smoking status Phenols: • Folic acid supplementation • BPA, BPS, BPF • Daily dietary caloric intake • Total bisphenols: BPA+BPS+BPF Infancia y Medio • 2004-2006 Phthalates: 2 urines • UPLC-MS/MS • Bottled water • 3-month FFQ (1st & 3rd • Creatinine adjusted chemical Ambiente • Spain • Sum-DEHP= • 1st trimester (Bioanalysis Research • Microwaved foods (plastic trimesters) • Maternal age (INMA) Project • N = 391 mEHP+mEHHP+mEOHP+mECPP+mCMHP • 3rd trimester Group at the Hospital containers) • Whole pregnancy • Country of origin (90) • mBzP del Mar Medical • Organic food questionnaire: organic • Pre-pregnancy BMI • mEP, miBP, mnBP Research Institute) • Milk, yogurt, cheese food, plastics • Education • Packed meat (sausages, pates) (3rd trimester) • Urine collection time • Canned fish • Potato chips • Canned beverages (soda, beer) • Other canned food (soups, sauces) Healthy Start • 2009-2014 Phthalates: 1 urine • HPLC-MS/MS (CDC) • Milk • 3-month food propensity • Creatinine adjusted chemical Pre-Birth Cohort • Colorado • Sum-LowMWPs = • 24-32 wks • Cheese questionnaire • Maternal age (119) • N = 446 mMP+mEP+miBP+mBP+mHiBP+mHBP • Yogurt (24-32 wks) • Race • Sum-DBP = mBP+miBP+mHBP+mHiBP • Ice cream • Questionnaire: fish oil • Pre-pregnancy BMI • Sum-HighMWPs = • Soft drinks supplement • Income mBzP+mEHP+mNP+mEOHP+mEHHP+ • Processed meat (24-32 wks) • Education mECPP+mCOP+mCNP • Red meat • Marital status • Sum-DEHP = mEHP+mEOHP+mEHHP+mECPP • Seafood • Employment Status Phenols: • Tofu • BPA, BPS • Fish oil supplements • Sum-parabens = methyl+ethyl+propyl+butyl • 2,4-dichlorophenol, 2,5-dichlorophenol • Triclosan, benzophenone-3 Puerto Rico • 2010-2012 Phthalates: 3 urines • HPLC-MS/MS (CDC) • Milk, cheese, ice cream • 48 hour questionnaires • Specific gravity adjusted Testsite for • Northern Puerto • mEHP, mEHHP, mEOHP, mECPP, mCOP, mCNP, • 18 wks • Meat, chicken, fish (18, 22, & 26 wks) chemical Exploring Rico mCPP, mBzP • 22 wks • Cold cuts, hot dog, sausage • Covariates that were Contamination • N = 139 • mEP, mnBP, miBP • 26 wks • Microwaved foods/drinks (plastic associated with each specific Threats container) chemical metabolite (PROTECT) (92) • Bottled water Australian • 2008-2011 Phenol: 1 urine • HPLC-MS/MS • Canned foods • Pregnancy questionnaire • None Maternal • Western BPA • 38 wks • (The National Centre • Microwaved foods/drinks (plastic • (38 wks) Exposure to Australia for Environmental container) Toxic • N = 26 Toxicology) • Plastic container storage Substances • Refillable bottles Study (AMETS) (285) 220 Table 18 (cont’d). Study Name Recruitment, Urine Urinary Chemical Chemicals Assessed Chemical Analysis Foods Assessed Dietary Assessment (Reference) Location, N Samples Adjustment & Covariates Infant • 2010-2012 Phthalates: 1 urine • HPLC-MS/MS • Peanut butter • “Typical week” FFQ (1st • Specific gravity adjusted Development • Minnesota, • Sum-DEHP = mEHP+mEHHP+mEOHP+mECPP 1st trimester (CDC) • Beef, poultry trimester) chemical and New York, • mBzP • HPLC-ESI-MS/MS • Other meats (pork, lamb) • General pregnancy • Maternal age Environmental Washington, • mEP • (Environmental • Oils and fats (butter, lard), Behavior/Lifestyle • BMI Study (TIDES) California • mBP Health Laboratory spices questionnaire • Race (282) • N = 656 miBP at the University of • Soy, dairy, fast food (1st trimester) • Study center • Washington) • Bottled beverages • Education • Organic/chemical free food • Grown/raised/caught food • Unprocessed food • Canned fruit or vegetables • Frozen fruit or vegetables Center for the • 1999-2000 Phenol: 2 urines • HPLC-MS/MS • Soda • Alcohol and soda • Specific gravity adjusted Health • California • BPA • 5-28 wks (CDC) • Alcohol consumption throughout chemical Assessment • N = 491 • 18-39 wks • Canned fruit pregnancy (5-28 & 18- • Maternal age of Mothers • Bottled water 39 wks) • Years in U.S. and Children • Pizza • Modified 3-month Block • Pre-pregnancy BMI of Salinas • Fish FFQ • Income/poverty ratio (CHAMACOS) • Hamburgers (18-39 wks) • Education (283) • Marital status • Parity • Urine collection time • Smoke exposure • Alcohol and soda intake Health • 2003-2006 Phenol: 3 urines • HPLC-MS/MS • Fresh or Frozen fish (store) • Frequency of • Creatinine adjusted Outcomes • Ohio BPA • 16 wks (CDC) • Fresh fruit or vegetables (store) consumption chemical and Measures • N = 389 • 26 wks • Canned fruit or vegetables questionnaire • Maternal age of the • Within 24 • Organic foods (conception to 20 wks & • Race Environment hrs of • Vegetarianism 20 wks to birth) • Income (HOME) delivery • Education Study (284) • Marital Status Sabadell Birth • 2004-2006 Phenol: 2 urines • LC-MS • Milk, yogurt • 3-month semi- • Creatinine adjusted Cohort • Spain BPA • 12 wks (The Department • Packaged meat (sausages, quantitative FFQ (12 & chemical (INMA • N = 479 • 32 wks of Analytical pates) 32 wks) • Maternal age Project) (281) Chemistry • Non-packaged meat (pork, • Pregnancy • Pre-pregnancy BMI Laboratory) chicken) questionnaire: water, • Social class • Canned fish organic foods, • Education • Non-canned fish (white fish, microwaving • Urine collection time seafood) foods/drinks • Smoking status • Potato chips (32 wks) • 2nd hand smoke exposure • Canned beverages (soda, beer) • Other canned foods (soups, sauces) • Fruits, Vegetables (fresh) • Bottled water (consumption & cooking) • Organic food • Microwaved foods/drinks (plastic container) BPA, bisphenol A; BPF, bisphenol F; BPS, bisphenol S; DEHP, di(2-ethylhexyl) phthalate; DOP, di-n-octyl phthalate; HPLC-EIS-MS/MS, HPLC-electrospray ionization-tandem MS; HPLC-MS/MS, HPLC-tandem MS; mBP, mono-n- butyl phthalate; mBzP, monobenzyl phthalate; mCMHP, mono(2-carboxymethyl)hexyl phthalate; mCNP, monocarboxynonyl phthalate; mCOP, monocarboxyoctyl phthalate; mCPP, mono(3-carboxypropyl) phthalate; mECPP, mono(2- ethyl-5-carboxypentyl) phthalate; mEHHP, mono(2-ethyl-5-hydroxyhexyl) phthalate; mEHP, mono(2-ethylhexyl) phthalate; mEOHP, mono(2-ethyl-5-oxohexyl) phthalate; mEP, monoethyl phthalate; mHBP, mono-hydroxybutyl phthalate; mHiBP, mono-hydroxyl-isobutyl phthalate; mHpP, mono-2-heptylphthalate; mHxP, mono-hexylphthalate; miBP, mono-isobutyl phthalate; mMP, mono-methylphthalate; mnBP, mono-n-butyl phthalate; mNP, mono-isononyl phthalate; sum-DBPs, sum of di-n-butyl phthalate metabolites; sum-DEHPs, sum of di(2-ethylhexyl) phthalate metabolites; sum-highMWPs, sum of high-molecular-weight phthalate metabolites; sum-lowMWPs, sum of low-molecular- weight phthalate metabolites; UPLC-MS/MS, ultra-performance LC-tandem MS. 221 Table 19. Dietary predictors of phthalate exposure in pregnant women Foods DEHP DiNP DiDP DOP BBzP HighMWPs Phthalic acid DEP DBP DiBP LowMWPs Packaged Meat NONE (90) NONE (90) NONE (90) NONE (90) NONE (90) NONE (92) NONE (92) NONE (92) (sausages) ↓ (92) NONE (92) NONE (92) NONE (92) NONE (92) Processed Meat NONE (119) NONE (119) NONE (119) NONE (119) Red meat NONE (119) NONE (119) NONE (282) NONE (119) NONE (282) NONE (282) NONE (119) Meat (beef, pork, lamb) NONE (282) NONE (282) NONE (92) NONE (92) NONE (92) NONE (92) NONE (92) Poultry/Chicken ↑ (92) NONE (92) NONE (92) NONE (282) NONE (282) NONE (282) NONE (282) NONE (282) Unspecified Meat NONE (92) NONE (92) NONE (92) NONE (92) NONE (92) NONE (92) NONE (92) NONE (92) Cold cuts, Hot Dog NONE (92) NONE (92) NONE (92) NONE (92) NONE (92) NONE (92) NONE (92) NONE (92) NONE (119) NONE (119) NONE (119) NONE (119) Seafood NONE (93) NONE (93) NONE (282) NONE (93) NONE (282) NONE (282) NONE (93) NONE (282) NONE (93) Seafood NONE (282) Canned Fish ↑ (90) NONE (90) NONE (90) NONE (90) NONE (90) Unspecified Fish NONE (92) NONE (92) NONE (92) ↓ (92) NONE (92) NONE (92) ↓ (92) NONE (92) Unspecified Vegetables ↑ (93) ↑ (93) ↑ (93) NONE (93) NONE (93) Fruits/ Canned Fruit & Vegetables NONE (282) NONE (282) NONE (282) NONE (282) NONE (282) Vegetables Frozen Fruit & Vegetables NONE (282) NONE (282) NONE (282) NONE (282) NONE (282) NONE (119) ↓ (119) ↑ (90) NONE (90) NONE (90) Milk NONE (90) NONE (92) NONE (92) NONE (92) NONE (119) ↑ (90) NONE (119) NONE (92) NONE (92) NONE (92) NONE (92) NONE (92) NONE (119) NONE (119) Yogurt NONE (90) NONE (119) NONE (90) NONE (90) NONE (119) NONE (90) ↓ (90) NONE (119) NONE (119) Dairy NONE (90) NONE (90) NONE (90) Cheese ↑ (90) NONE (92) NONE (92) NONE (92) NONE (119) NONE (90) NONE (119) Products ↓ (92) NONE (92) NONE (92) NONE (92) NONE (92) NONE (119) NONE (119) Ice Cream NONE (92) ↑ (92) ↑ (92) NONE (92) NONE (119) NONE (92) NONE (92) NONE (119) NONE (92) NONE (92) Unspecified Dairy NONE (282) NONE (282) ↑ (282) NONE (282) ↓ (282) Oils, Butter, Lard, NONE (282) NONE (282) NONE (282) NONE (282) NONE (282) Shortenings Fast Food NONE (282) NONE (282) NONE (282) NONE (282) NONE (282) Fast Food Potato Chips NONE (90) NONE (90) NONE (90) NONE (90) NONE (90) 222 Table 19 (cont’d). Foods DEHP DiNP DiDP DOP BBzP HighMWPs Phthalic acid DEP DBP DiBP LowMWPs Organic, NONE (90) ↓ (90) NONE (90) NONE (90) NONE (90) Organic Food Chemical-Free, NONE (282) NONE (282) ↓ (282) NONE (282) NONE (282) and “Marked” Organic Foods ↑ (282) NONE (282) NONE (282) NONE (282) NONE (282) Environmentally Grown/raised/caught food NONE (282) NONE (282) ↓ (282) NONE (282) ↓ (282) Friendly Foods Unprocessed Food NONE (282) NONE (282) NONE (282) NONE (282) NONE (282) Soups & Bouillon NONE (93) NONE (93) NONE (93) NONE (93) ↓ (93) Spices ↓ (282) ↓ (282) NONE (282) ↓ (282) ↓ (282) Other Foods and Peanut Butter NONE (282) NONE (282) NONE (282) NONE (282) NONE (282) Dietary Patterns Grains NONE (93) NONE (93) NONE (93) ↓ (93) NONE (93) Daily Dietary Caloric Intake NONE (93) NONE (93) NONE (93) NONE (93) NONE (93) (< & > 2000-2399 kcal) Tofu NONE (119) NONE (119) NONE (119) NONE (119) Soy Soy NONE (282) NONE (282) NONE (282) NONE (282) NONE (282) Fish Oil Supplements NONE (119) NONE (119) NONE (119) NONE (119) Supplements Folic Acid Supplements ↓ (93) ↓ (93) ↓ (93) NONE (93) NONE (93) Food/Water Stored in NONE (92) NONE (92) NONE (92) NONE (92) NONE (92) NONE (92) NONE (92) NONE (92) Plastic Containers ↑ (137) ↑ (137) NONE (137) ↑ (137) NONE (137) NONE (90) NONE (90) NONE (90) NONE (90) NONE (90) Plastic Bottled Water/Other Drinks NONE (92) NONE (92) NONE (92) NONE (92) NONE (92) ↑ (92) NONE (92) NONE (92) Containers and NONE (282) NONE (282) NONE (282) NONE (282) NONE (282) Tableware NONE (90) NONE (90) NONE (90) NONE (90) Foods/Drinks Microwaved NONE (90) NONE (92) ↑ (92) NONE (92) NONE (92) NONE (92) NONE (92) NONE (92) in Plastic Containers NONE (92) ↑ (137) ↑ (137) NONE (137) ↑ (137) NONE (137) Plastic Tableware NONE (137) ↑ (137) NONE (137) NONE (137) NONE (137) NONE (119) Canned Beverages NONE (119) NONE (119) NONE (119) NONE (93) NONE (93) NONE (90) NONE (93) NONE (90) NONE (90) (soft drinks, beer) NONE (93) NONE (90) ↑ (93) Canned Products NONE (90) Other Canned Food (soups, NONE (90) ↑ (90) NONE (90) NONE (90) NONE (90) sauces) References are provided in parentheses. ↑: positive association; ↓: negative association; NONE: no association; Empty fields: associations not assessed between food product and urinary phthalate metabolite concentrations. HighMWPs: high-molecular weight phthalates; LowMWPs: low-molecular weight phthalates. *Phthalate exposure was predicted from phthalate parent-specific metabolites in urine. Results for some parent phthalates (DEHP, DBP) represent either metabolite-specific or molar sum-metabolite associations. 223 Table 20. Dietary predictors of bisphenol exposure in pregnant women Foods Bisphenol A Bisphenol S Bisphenol F Total Bisphenols Packaged Meat (sausages) NONE (281) Non-Packaged Meat (chicken, pork) NONE (281) Meat Processed Meat NONE (119) NONE (119) Red Meat NONE (119) NONE (119) NONE (119) NONE (119) Seafood NONE (93) ↑ (93) ↑ (93) NONE (93) NONE (281) Seafood Fresh, Frozen, Un-canned Fish NONE (284) Canned Fish ↑ (281) Unspecified Fish NONE (283) NONE (281) Fresh/Non-Canned Fruits NONE (284) NONE (283) Canned Fruits NONE (284) Fruit and NONE (281) Vegetables Fresh/Non-Canned Vegetables NONE (284) Canned Vegetables ↑ (284) Unspecified Vegetables NONE (93) NONE (93) NONE (93) NONE (93) Organic Fruit and Vegetables NONE (284) NONE (119) Milk NONE (119) NONE (281) NONE (119) Dairy Products Yogurt NONE (119) NONE (281) Cheese NONE (119) NONE (119) Ice Cream NONE (119) NONE (119) Hamburgers ↑ (283) Fast Food Pizza NONE (283) Potato Chips NONE (281) Organic and NONE (281) Organic Food Environmentally NONE (284) Friendly Foods Vegetarianism ↓ (284) Alcohol NONE (283) Soups & Bouillon NONE (93) NONE (93) NONE (93) NONE (93) Other Foods and Tofu NONE (119) NONE (119) Dietary Patterns Grains NONE (93) ↓ (93) NONE (93) NONE (93) Daily Dietary Caloric Intake NONE (93) ↓ (93) NONE (93) NONE (93) (< & > 2000-2399 kcal) Fish Oil Supplements NONE (119) NONE (119) Supplements Folic Acid Supplements ↓ (93) NONE (93) NONE (93) NONE (93) Foods/Drinks Microwaved in Plastic NONE (281) Containers NONE (285) NONE (281) Plastic Containers Bottled Water NONE (283) NONE (285) Foods Stored in Plastic Containers NONE (285) NONE (119) Canned Beverages NONE (93) NONE (119) NONE (93) NONE (93) (soft drinks, beer) NONE (281) NONE (93) Canned Products ↑ (283) NONE (281) Canned Foods NONE (285) References are provided in parentheses. ↑: positive association; ↓: negative association; NONE: no association; Empty fields: associations not assessed between food product and urinary bisphenol concentrations. *Bisphenol exposure was predicted from their concentrations in urine. Total bisphenols include 3 bisphenols (BPA, BPS, and BPF) measured in urine. 224 A.5. CURRENT STATUS OF KNOWLEDGE Food-monitoring studies are performed worldwide to evaluate the safety of foods and dietary patterns, including assessing exposures to environmental chemicals through certain dietary practices. These studies are performed by government agencies, as well as independent laboratories around the world. For example, the U.S. FDA has ongoing food-monitoring programs such as the Total Diet Study (TDS) and the Chemical Contaminants Monitoring Program to examine the safety of foods on the U.S. market (288). Through these programs, the FDA collects information about consumer food preparation and consumption practices (TDS), as well as the potential exposure to and risk of chemical contaminants found in the U.S. food supply (TDS and Chemical Contaminants Monitoring Program) (288). Similarly, the European Food Safety Authority carries out risk assessments and food-monitoring studies within the European Union to determine the safety of chemical contaminants in foods consumed by humans and animals (289). However, many of these government food-monitoring programs have not assessed concentrations of phthalates and bisphenols in foods. Therefore, the food- monitoring studies assessing exposure to phthalates and bisphenols from the food supply (reviewed below) were conducted by independent laboratories around the world. The goal of these food-monitoring studies was to evaluate the potential for human exposure to phthalates and bisphenols through food by capturing dietary habits from various countries, including the U.S., Canada, United Kingdom, France, Spain, Norway, Belgium, Tunisia, Israel, China, and Japan. There are extensive reviews in the literature summarizing food-monitoring studies that measured phthalate and BPA concentrations 225 in foods (239, 290), and some of the results from these reviews and additional food- monitoring studies are summarized below. A.5.1. Food-monitoring studies A.5.1.1. Meat Foods of animal origin, including beef, pork, and poultry, are major sources of HighMWPs and BPA from processing and packaging (291), partially because HighMWPs, and to a lesser extent LowMWPs, are slightly lipophilic and can bioaccumulate in fat-containing foods (292). International food-monitoring studies consistently report high detectable concentrations of HighMWPs (especially DEHP) and BPA in meat and meat products (239, 293-298). Food-monitoring studies have also reported low, but detectable concentrations of LowMWPs (compared to HighMWPs (239)), in meat and meat products, suggesting that meats may also be important sources of DEP, DBP, and DiBP (293, 294). A.5.1.2. Seafood Numerous food-monitoring studies from the United Kingdom, Norway, Belgium, China, and the U.S. have reported detectable concentrations of phthalates and BPA in seafood products (293-295, 299). Similar to other foods packaged in plastics (257, 293) and cans (298, 300), these food-monitoring studies also suggest that food packaging materials contribute to phthalate and BPA concentrations detected in seafood products. A study in Spain found that 34.7% of Spanish pregnant women reported consuming canned fish > 1-3 times/week, making it the most frequently-consumed canned food in this population, and a major source of BPA during pregnancy (301). 226 A.5.1.3. Fruits and vegetables Fruit and vegetable consumption is considered a measure of healthier lifestyles associated with lower EDC exposures (298, 302). Food-monitoring studies from Belgium, France, China, and the U.S. report low concentrations of both HighMWPs and LowMWPs in fruit and vegetable products, suggesting low likelihood for phthalate exposure from these foods (294-296, 303, 304). However, one study has shown that exposure to HighMWPs and LowMWPs in vegetables primarily comes from ready-to-eat vegetables (e.g. lettuce, arugula, parsley, carrot, and corn salad) packaged in plastic bags (305). BPA food-monitoring studies from Norway, Canada, and the U.S. suggest that canned fruits and vegetables, rather than fresh, are major exposure sources (298, 306), and that overall concentrations of BPA in noncanned fruits and vegetables are relatively low (293, 297, 307). A.5.1.4. Dairy products Milk, yogurt, cheese, ice cream, and butter can be high in fats, making it possible for phthalates to accumulate in these foods (292). Analyses of milk in Belgium found higher levels of HighMWPs (DEHP and BBzP) and LowMWPs (DBP and DiBP) in milk retail products compared to raw cow’s milk, suggesting that phthalates can migrate into raw cow’s milk from contaminated feed ingested by cows, during the mechanical milking process, and/or from milk packaging materials used at the dairy factory (308). Food- monitoring studies reported detectable concentrations of DEHP in select cheese samples from Canada (254), DEHP/DiNP/DOP in milk and cheese from Norway (293), DEHP/DOP/BBzP in milk and other dairy from the U.S. (295), DEHP in milk and dairy 227 products from Belgium (294), and BBzP in milk, butter/oil, and yogurt from Tunisia (309). In contrast, Norwegian and the U.S. food-monitoring studies reported low or undetectable concentrations of LowMWPs in milk and milk products (239, 293-295). Similar to HighMWPs, BPA has also been detected in milk and dairy products. In Europe, higher BPA levels were found in canned milk and dairy products compared to un-canned products (European Food Safety Authority) (310). The BPA concentrations in European dairy are consistent with those from China (311), Canada (298), and Japan (312). Detectable BPA concentrations were also found in dairy products (milk, ultra-fresh dairy products, and cheese) consumed by French pregnant women (313), suggesting that these are important sources of BPA exposure in pregnancy. A.5.1.5. Fast food Food-monitoring and epidemiological studies suggest that fast food and/or foods served at restaurants are likely sources of phthalate and BPA exposures (239, 292). Although phthalates and bisphenols could leach into foods during processing, they have also been shown to migrate into foods from packaging, including pizza boxes or sandwich wrappers (314). A food-monitoring study in Canada reported detectable concentrations of BPA in some fast food products (French fries, hamburgers, and sandwiches), but not others (pizza and chicken nuggets) and concluded that hamburgers had the highest BPA concentrations of all fast foods assessed (298). An assessment of fast food intake in the general U.S. population (NHANES 2005-2014) found that compared with nonconsumers, both low and high consumers of food away from home had significantly higher 228 concentrations of urinary sum of di(2-ethylhexyl) phthalate (Sum-DEHP) and monocarboxyoctyl phthalate [but not monobenzyl phthalate (mBzP), mEP, mono-n-butyl phthalate (mBP), or mono-isobutyl phthalate (miBP) (315)]. A similar study further analyzed fast food consumption in the U.S. population by food group, and found that fast food grain, including pizza, intake was associated with urinary Sum-DEHP and sum of di- isononyl phthalate metabolites (Sum-DiNP), whereas potato chip and hamburger intakes were associated with higher urinary Sum-DEHP and Sum-DiNP (316). A.5.1.6. Organic & environmentally-friendly food and dietary patterns Organic, chemical-free and environmentally friendly foods are generally considered markers of healthier lifestyles, but associations of these dietary patterns with EDC exposures are not well-characterized. Although the literature is limited, a dietary intervention that focused on exclusive consumption of fresh and organic foods for three days found that urinary DEHP metabolite concentrations decreased by 53-56% and urinary BPA concentrations decreased by 66% lower from pre- to during-intervention. However, foods provided to and prepared by these participants were from plastic-free packaging and nonplastic containers (291), making it difficult to determine if the decreases in urinary chemical concentrations were due to the foods’ organic status or their packaging/preparation materials. Similarly, residents of a rural vegetarian/vegan community in Israel had significantly lower urinary phthalate metabolite concentrations, but BPA concentrations in these individuals were not different from those in the general Israeli population (317). However, a U.S. dietary intervention with fresh and organic foods prepared without plastics found significantly higher urinary DEHP metabolites after the 229 intervention, which was due to increased intake of certain foods (e.g., spices and peanut butter) (318). A.5.1.7. Plastic containers and tableware Both phthalates and BPA are used to manufacture plastics for food storage and cooking containers (256). Data suggest that these are important sources of human EDC exposure, as these EDCs can migrate from plastic containers and tableware into foods and beverages, especially during heating and cooling (239, 301). For phthalates, migration levels of DBP were higher with prolonged plastic container use and longer heating time (319), and high levels of DEHP and DBP were found in plastic tableware at room temperature (320). BPA, however, migrates from reusable polycarbonate plastic water bottles into water at room and high temperatures (321), and the use of polycarbonate water bottles has been shown to increase urinary BPA concentrations (322), especially during the hot summer months (323). A.5.1.8. Canned foods and beverages BPA is used to manufacture polycarbonate and epoxy resins for metal can linings and is detectable in a variety of canned foods (293, 298, 324-326). A study characterizing dietary BPA exposure in the French population (including pregnant women) identified that canned products accounted for ~50% of total BPA exposure (313). Phthalates are generally found in plastic food packaging materials, so their concentrations in canned products are low, but detectable (293). 230 A.5.2. Dietary Predictors of Phthalates and Bisphenols in Pregnancy Figure 14 summarizes potential sources of phthalates and bisphenols from food packaging materials and consumer food practices. As previously mentioned, both diet and physiology are greatly modified in pregnancy (277-280). Thus, despite the many food-monitoring studies assessing dietary sources of phthalates and BPA in the general public, studies specifically in pregnancy are important for establishing dietary recommendations for this vulnerable population. In the 10 studies reviewed here, urinary concentrations of phthalates and bisphenols in women from U.S. studies were comparable to those of women in the general U.S. population (using data from the 2013- 2014 NHANES (327)). The studies performed outside the U.S. recruited cohorts to reflect their individual populations, suggesting that exposures described in these studies likely represent the general population of each country. Overall, in pregnancy, associations between use of plastic containers and increased urinary phthalate metabolite concentrations, and between consumption of canned foods and increased urinary BPA concentrations were consistent with previous food-monitoring studies. Foods and dietary patterns associated with a healthier lifestyle, such as organic foods, grown/raised/caught foods, vegetarianism, and folic acid supplementation, as well as some other dietary patterns and foods, including soups and bouillon, spices, and grains, were generally associated with lower urinary phthalate metabolite and bisphenol concentrations in pregnant women. However, despite the many food-monitoring studies reporting high phthalate and BPA concentrations in meats and dairy, the designs of most studies in pregnant women were unable to reliably detect associations of specific foods/food groups 231 with phthalates and BPA. Some of the differences and inconsistencies in study design are highlighted in Table 18 and will be discussed in the following sections. Figure 14. Summary of potential food packaging materials and consumer food practices. (A) Potential determinants of phthalate and BPA exposures through food packaging materials, including soft and hard plastics, cans, lined paper, and glass. (B) Consumer food practices, including food preparation, as well as storage and consumption materials that influence exposure to phthalates and BPA. BPA, bisphenol A. A.5.2.1. Measuring urinary phthalates and bisphenols Some technical/analytical factors may be important to consider when evaluating currently available studies in pregnant women. First, the analytical methodologies vary slightly among the 10 studies reviewed here (Table 18), and additional studies may be needed to determine whether unifying these methods would eliminate reported inconsistencies between studies. Another important technical consideration when using urine to assess chemical exposures is establishing appropriate methods to account for urine dilution across participants. In the studies reviewed here, two studies did not adjust for urinary dilution, whereas five adjusted for creatinine, and three adjusted for specific gravity (Table 232 18). Urinary density can vary greatly across pregnancy and depending on hydration status, therefore raw values of phthalate and bisphenol concentrations need to be adjusted for some measure of urine density. Furthermore, it has been suggested that creatinine in pregnancy is affected by pregnancy-related fluid dilution, making specific gravity adjustment a more accurate approach in pregnant women (328). A.5.2.2. Evaluating exposure to phthalates and bisphenols Most studies in pregnant women reviewed here included analyses of phthalates and bisphenols, but classification of these chemicals often varied across studies. For example, findings for associations between milk consumption and urinary mBP (metabolite of DBP) in pregnancy were conflicting. One study reported that the sum of di- n-butyl phthalate metabolite concentrations [composed of mBP, miBP, mono- hydroxybutyl phthalate (mHBP), and mono-hydroxyl-isobutyl phthalate (mHiBP)] was negatively associated with milk consumption (119), whereas other studies found no association (92) or a positive association (90) between milk consumption and urinary mBP, and no association with urinary miBP (90, 92). Similar examples were also observed for sum or individual measures of other phthalates and bisphenols (Tables 18 – 20), which suggests that classifying chemicals into their predicted exposure sources (parent chemicals) may yield differing results compared with assessment of individual breakdown metabolites. Studies in pregnant women report low reproducibility and sensitivity of urinary phthalate metabolites and bisphenols across pregnancy (18, 108, 329-331). To overcome this 233 challenge, several of the studies reviewed here collected two (90, 281, 283) or even three (92, 284) urine samples across pregnancy to evaluate gestational exposure to bisphenols and phthalates. Assessing these relations at multiple times in pregnancy may be critical to account for some of the physiological and dietary shifts that occur in early compared with late pregnancy. However, aligning the timing of exposure assessment in relation to diet may be more important to accurately and consistently predict dietary sources of chemical exposure. Based on the current review of the literature, only one study accurately captured dietary sources of phthalates and bisphenols (92), as it assessed chemical exposure within 48 hours of the dietary report, whereas other studies used surveys that assessed much broader windows of dietary intakes. This appears to be a critical difference in study design, as many observations differed between this study and others. For example, in one study, milk consumption assessed by a 3-month FFQ was associated with increased urinary mBzP concentrations (90), whereas mBzP assessed within 48 hours of dietary assessment was not associated with milk consumption (92). Similarly, higher cheese consumption was associated with lower urinary mBzP (92) when assessed within 48 hours of the dietary survey, whereas no associations were reported with urinary mBzP in a study assessing diet using a 3-month dietary questionnaire (90). Given that the study assessing chemical exposure within 48 hours of the dietary report was limited to women in Northern Puerto Rico, future studies in other pregnant populations should expand on findings from this study by assessing chemical exposure at multiple times across pregnancy, and within 24-48 hours of dietary assessment. 234 A.5.2.3. Assessing diet in pregnancy Dietary patterns are most-often assessed using FFQs, which are inexpensive, easy to use, and validated to reflect long-term dietary intake patterns in pregnant populations (332, 333). Because FFQs and other similar questionnaires are designed to assess dietary intake of nutrients, not chemicals in food, they may not accurately predict dietary sources of phthalate and BPA exposure. Especially problematic is that information collected from these questionnaires spans weeks or months, whereas the short half-lives of EDCs mean that urinary exposure assessment reflects their concentrations within (often) 24-hours of assessment. Based on the aforementioned examples, assessment of diet within ~24-hours of exposure assessment will be the most appropriate approach for quantifying dietary sources of phthalates, bisphenols, and other environmental chemicals that have relatively short half-lives. The other critical factor to consider when evaluating currently available studies in pregnancy is the broad range of dietary surveys used in these studies (Table 18). First, many studies utilized exceedingly general categories of various food types. Examples of this include: using the vague category of “processed meat” (119), the term “seafood” (93, 119, 282) to refer to all types of fish or shellfish, the combined analysis of all “fruits and vegetables” (282), and the overly broad analysis of “fast food” (282) without specifying the food item or restaurant category. These general categories make it difficult to establish patterns of chemical exposures from specific foods, and to ultimately provide pregnant women with specific recommendations as they make food-purchasing decisions. 235 More importantly, many studies reviewed here acknowledged that their dietary surveys were unable to distinguish between various categories of food packaging, including fresh, frozen, or packaged foods. For example, in the Infancia y Medio Ambiente Project, first- and second-trimester consumption of canned fish was associated with higher urinary BPA concentrations (281), whereas two other studies found no associations between fresh, frozen, or un-canned fish consumption and urinary BPA (281, 284). BPA is used to manufacture polycarbonate and epoxy resins for metal can linings and is detectable in a variety of canned foods (293, 298, 324-326), so observations of higher BPA concentrations with consumption of canned (but not other) fish are consistent with food- monitoring studies. However, several other studies assessed seafood or fish intake with BPA, but the dietary questionnaires in these studies did not distinguish between fresh or canned fish, making it difficult to conclusively interpret these findings (93, 119, 283). Future studies should utilize dietary questionnaires that robustly assess both diet and the mode of food packaging/preparation by asking specifically about how each food item was packaged (e.g., fresh, canned, plastic, glass) and prepared (e.g., microwaved, steamed in plastic or glass). A.5.2.4. Considering demographic and lifestyle factors Another major challenge for establishing dietary correlates of EDC exposure are the cultural differences in both dietary intake and chemical production/food packaging. For example, associations of meat and dairy intake with phthalates and bisphenols differed greatly between pregnancy studies in the contiguous U.S., Puerto Rico, and Spain. These differences likely reflect variability in the types and amounts of meats/dairy foods 236 consumed in these cultures, and how these foods are processed/packaged. Although it is important to establish generalizable recommendations for pregnant women, these cultural differences may require culturally-sensitive recommendations. Most studies reviewed here acknowledged that many of the unexpected associations observed between some particular EDCs and foods may be due to other concomitant lifestyle factors. Here, and in other pregnancy cohorts, these factors included maternal age, race/ethnicity, country of origin, pre-pregnancy BMI, marital status, education, employment status, annual income, and personal care/household product use (18, 90, 92, 93, 108, 119, 124, 281-284, 330, 331). For example, mEP concentrations were higher in pregnant women who used bottled water for cooking 48 hours before exposure assessment than in women who used the public water supply (92). Given that DEP (mEP’s parent compound) is primarily found in personal care products, the authors of this study suggest that the positive association between bottled water use and mEP is unrelated to use of plastics, and could be explained by higher urinary mEP concentrations in women reporting use of perfume/cologne and colored cosmetics (92). In addition to carefully controlling for other important lifestyle factors in statistical models, if dietary questionnaires fail to capture cultural or geographic differences in food packaging or processing, future studies may need to investigate these factors as modifiers of food- exposure relations. For example, annual income or employment status may affect seafood choices. Therefore, although no overall associations may be observed between fish consumption and phthalate exposure, positive associations are possible in low- 237 income groups that tend to consume canned or packaged fish, but not in higher-income women who tend to eat fresh or unpackaged fish. A.6. CONCLUSION Maternal diet is an established determinant of a healthy pregnancy and fetal outcomes. Phthalates and BPA are known to affect pregnancy and fetal development, and the 10 studies reviewed here suggest that certain dietary components or patterns are important sources of BPA, HighMWPs, and LowMWPs in pregnancy. However, consistencies in observed associations between studies were limited to long-term lifestyle choices, including those related to home food preparation (use of plastics to store, prepare, and heat foods) and choices of food types (canned, organic). Several well-designed studies in non-pregnant populations do suggest that changing dietary behaviors can limit exposure to phthalates and bisphenols. For example, a dietary intervention of five families in the U.S. (n=20) found significant reductions in urinary BPA and DEHP metabolites after limited consumption of foods packaged and prepared in plastics and cans, and increased concentrations of these chemicals with resumption of packaged food consumption (291). Furthermore, a strict 48-hour fasting study of five individuals from Germany observed significant reductions in urinary HighMWPs from pre-fast to post-fast, whereas urinary LowMWP concentrations stayed consistent throughout the study (128). These findings suggest that addressing the inconsistencies in study designs among the pregnancy studies described here could provide valuable insight for establishing specific EDC- limiting dietary recommendations to improve pregnancy and fetal outcomes. 238 APPENDIX B: MATERNAL DIET QUALITY MODERATES ASSOCIATIONS BETWEEN PARABENS AND BIRTH OUTCOMES This article/appendix has been published in Environmental Research; Volume 214 Part 3; Pacyga DC, Talge NM, Gardiner JC, Calafat AM, Schantz SL, Strakovsky RS; Maternal diet quality moderates associations between parabens and birth outcomes. Copyright Elsevier (2022); https://doi.org/10.1016/j.envres.2022.114078. B.1. ABSTRACT Maternal paraben exposure and diet quality are both independently associated with birth outcomes, but whether these interact is unknown. We assessed sex-specific associations of parabens with birth outcomes and differences by maternal diet quality. Illinois pregnant women (n = 458) provided five first-morning urines collected at 8–40 weeks gestation, which we pooled for quantification of ethylparaben, methylparaben, and propylparaben concentrations. We collected/measured gestational age at delivery, birth weight, body length, and head circumference within 24 h of birth, and calculated sex-specific birth weight-for-gestational-age z-scores and weight/length ratio. Women completed three- month food frequency questionnaires in early and mid-to-late pregnancy, which we used to calculate the Alternative Healthy Eating Index (AHEI)-2010. Linear regression models evaluated sex-specific associations of parabens with birth outcomes, and differences in associations by average pregnancy AHEI-2010. In this predominately non-Hispanic white, college-educated sample, maternal urinary paraben concentrations were only modestly inversely associated with head circumference and gestational length. However, methylparaben and propylparaben were inversely associated with birth weight, birth 239 weight z-scores, body length, and weight/length ratio in female, but not male newborns. For example, each 2-fold increase in methylparaben concentrations was associated with −46.61 g (95% CI: −74.70, −18.51) lower birth weight, −0.09 (95% CI: −0.15, −0.03) lower birth weight z-scores, −0.21 cm (95% CI: −0.34, −0.07) shorter body length, and −0.64 g/cm (95% CI: −1.10, −0.19) smaller weight/length ratio in females. These inverse associations were more prominent in females of mothers with poorer diets (AHEI-2010 < median), but attenuated in those with healthier diets (AHEI-2010 ≥ median). In newborn males of mothers with healthier diets, moderate inverse associations emerged for propylparaben with gestational length and head circumference. Maternal diet may moderate associations of parabens with birth size in a sex-specific manner. Additional studies may consider understanding the inflammatory and metabolic mechanisms underlying these findings. B.2. KEYWORDS Parabens; birth size; gestational length; head circumference; newborn sex; maternal diet. B.3. INTRODUCTION A newborn’s gestational age at birth, birth size (weight and length), and head circumference are known predictors of neonatal morbidity and mortality (334), and are associated with maternal and child lifelong health. Babies born pre-term (birth before 37 weeks gestation) and early-term (birth after 37 weeks, but before 39 weeks gestation) have higher risks of developing metabolic disorders, respiratory issues, and cognitive 240 problems compared to full-term infants (birth after 39 weeks, but before 41 weeks) (335- 338). Similarly, late-term birth (birth after 41 weeks gestation) is associated with higher risk of delivery complications for both mother and newborn, as well as childhood and adulthood obesity (221, 339, 340). After accounting for gestational age at birth, smaller body size at birth also predicts child cognitive problems and other adverse developmental outcomes that may persist through adolescence (220). Similarly, larger birth size is associated with long-term cardiometabolic consequences for children, including increased risk of type 2 diabetes and obesity (221), but also with delivery complications that can adversely impact both mother and newborn (341). Finally, head circumference measured at birth has been shown to be a clinical indicator of newborn brain volume (342), though the consequences of this are not fully understood. Therefore, it is important to understand risk factors in pregnancy that influence a baby’s gestational age at birth, birth size, and head circumference. Greater than 90% of reproductive-aged women in the United States (U.S.) have measurable urinary biomarker concentrations of parabens (alkyl esters of p- hydroxybenzoic acid) (343). Parabens are anti-microbial agents found in many personal care products and cosmetics, as well as some food products and medications (344). Studies in cells, animal models, and pregnant populations suggest that parabens target inflammatory (345-349), hormonal (265, 350-353), and metabolic (354-356) pathways, which are important pathways for fetal growth. Additionally, observational studies have detected parabens in cord blood (357), amniotic fluid (258), and placental tissue (358), suggesting that parabens can cross the placenta. Consequently, some studies evaluating 241 associations of parabens with gestational length, birth size, and head circumference observed that parabens are associated with unfavorable birth outcomes (359-364). However, a 2019 systematic review qualitatively summarized six previous studies evaluating associations of paraben biomarkers with birth weight, length, and head circumference, and concluded that the strength and direction of associations are inconsistent across studies (365). Similarly, a 2020 meta-analysis pooled six to eight studies (depending on the paraben) evaluating associations of paraben biomarkers with birth weight, but observed no significant pooled associations for any paraben biomarker (366). More recently, several studies suggest that associations of parabens with birth size may be sex-specific (134, 360-362, 367, 368), though results from these studies have also been inconsistent. Therefore, further research can help clarify sex-specific associations of parabens with birth outcomes. Appropriate maternal diet is critical for optimal fetal growth and development, whereas poor maternal diet quality is associated with adverse birth outcomes, including shorter gestational length and smaller birth size (241, 369). Importantly, a growing body of literature suggests that healthy maternal diets may protect the developing fetus from the adverse effects of environmental exposures (370-372). With regards to birth size, some studies observed that adverse associations of maternal exposure to chemicals (e.g., per- and polyfluoroalkyl substances and metals) with birth weight were attenuated in newborns of mothers with better diets, as measured by consumption of select whole foods or specific individual macro- or micronutrients (373, 374). Studies suggest both parabens and diet may interact with similar inflammatory and metabolic pathways (244, 347, 375), 242 which are important mechanisms involved in regulating gestational length and fetal growth. However, no studies have evaluated diet as a moderator of associations between parabens and birth outcomes. Additionally, most prior studies evaluating diet as a moderator of associations between environmental chemicals and health outcomes only focused on individual nutrients or foods, while few considered assessing dietary patterns or diet quality (376). Utilizing diet indices that reflect diet quality accounts for the combined effect of many dietary behaviors and the interactive effects of foods and nutrients on health (102). For example, the Alternative Healthy Eating Index 2010 (AHEI-2010) focuses on foods predictive of chronic disease risk, and a higher AHEI-2010 score in pregnancy, specifically, has been associated with improved maternal and offspring outcomes (377-380). Additionally, higher AHEI-2010 scores are correlated with lower inflammation (381). Therefore, it may be important to consider holistic dietary patterns when evaluating whether diet can attenuate the negative impacts of environmental toxicant exposures, especially exposures to parabens. Given the current status of the literature, our first objective was to evaluate sex-specific associations of maternal urinary paraben concentrations with gestational length, birth size (birth weight, birth weight z-scores, body length, and weight/length ratio), and head circumference. Our second objective was to determine if sex-specific associations of paraben biomarkers with birth outcomes differ by maternal diet quality, which we evaluated using the AHEI-2010. 243 B.4. MATERIALS AND METHODS B.4.1. Illinois Kids Development Study (I-KIDS) recruitment and enrollment The current study includes pregnant women from I-KIDS, an ongoing prospective pregnancy/birth cohort designed to evaluate the impacts of prenatal environmental chemical exposures on infant neurodevelopment. I-KIDS recruitment and enrollment have been described elsewhere (86). Briefly, pregnant women were recruited at their first prenatal care appointment from two local obstetric clinics in Champaign-Urbana, IL. Women were eligible to participate if they were < 15 weeks pregnant, 18-40 years old, fluent in English, in a low-risk singleton pregnancy, living within a 30-minute drive of the University of Illinois campus, and not planning to move out of the area before their child’s first birthday. A total of 482 pregnant women enrolled in I-KIDS between 2013 and 2018 and remained in the study through the birth of their infant. We excluded the limited number of women who delivered before 37 weeks gestation, and those who were missing information on important covariates. Therefore, the current analysis includes a total of 458 mothers-newborn pairs (233 females and 225 males). These women provided written informed consent, and the study was approved by the Institutional Review Board at the University of Illinois. The analysis of de-identified specimens at the Centers for Disease Control and Prevention (CDC) laboratory was determined not to constitute engagement in human subjects’ research. B.4.2. Collection of maternal sociodemographic, lifestyle, and health information Immediately after enrollment, study staff visited I-KIDS participants at home to obtain information about their sociodemographic, lifestyle, and health characteristics. We 244 collected information about the following characteristics from an interviewer-administered questionnaire: age, race/ethnicity, pre-pregnancy weight, height, and parity. Self-reported pre-pregnancy weight and height were used to calculate pre-pregnancy BMI (kg/m2) (98- 100). To calculate conception season, we used reported due dates based on the first day of the last menstrual period, which we confirmed after the first trimester ultrasound. To assess early pregnancy depression symptoms, participants also completed an adapted version of the Edinburgh Postnatal Depression Scale (EPDS), which excluded the question “The thought of harming myself has occurred to me” (382). B.4.3. Assessment of urinary paraben biomarker concentrations Women provided at least three and up to five first-morning urine samples at study home visits or routine prenatal care visits at the following gestational timepoints: 8-15, 13-22, 19-28, 25-33, 32-40 weeks gestation (median 13, 17, 23, 28, and 34 weeks gestation, respectively) as described previously (86). Greater than 95% of women contributed all five urine samples, and the remaining 5% contributed three or four urines. Urine samples were collected in polypropylene urine cups and refrigerated immediately. Within 24 hours of collection, we aliquoted urine samples for long-term storage (at -80 °C) or pooled from each timepoint. To create the pool sample, we added 900 µL of fresh urine from each timepoint to a 5 mL cryovial, and the sample was immediately stored at -80 °C between each gestational timepoint. After the first visit, we layered fresh urine onto frozen urine from previous gestational timepoints (frozen urine was not thawed). At the end of pregnancy, we thawed, vortexed, and measured specific gravity of pooled samples. We shipped frozen pooled samples on dry ice to the CDC Division of Laboratory Sciences in 245 three batches, in the chronological order of enrollment (batch one enrolled December 2013 - February 2015, batch two enrolled February 2015 - July 2016, and batch three enrolled July 2016 - August 2018). Pooled urine samples were analyzed for urinary concentrations of the following four commonly measured paraben biomarkers: butylparaben, propylparaben, ethylparaben, and methylparaben using automated on-line solid phase extraction-high performance liquid chromatography–isotope dilution tandem mass spectrometry based on previously published methods (383, 384). These methods have rigorous quality control/quality assurance protocols and excellent long-term reproducibility, and inter- and intra-day variability were < 9% indicating good precision for all measured paraben concentrations (384). The limits of detection were 0.1 ng/mL for butylparaben and propylparaben and 1.0 ng/mL for ethylparaben and methylparaben. B.4.4. Collection of gestational length, birth size, and head circumference measures Within 24 hours of birth, I-KIDS staff visited participants at the hospital to obtain delivery date, which we used to calculate gestational length, and to conduct newborn anthropometric measurements. Hospital staff measured birth weight (g) immediately after delivery, which we obtained from crib cards. Sex-specific birth weight for gestational age z-scores were calculated using a U.S. population-based reference according to published methods (385). I-KIDS staff measured body length (cm; Seca Light & Stable Measuring Board) and head circumference (cm; flexible retractable ruler) in triplicate, and the mean of multiple values was used in statistical analyses. In the few cases where we were unable to measure body length and head circumference at the hospital (n = 28 and 9, 246 respectively), we used body length and head circumference information obtained from hospital crib cards. We calculated newborn weight/length ratio (g/cm) by dividing birth weight by body length. B.4.5. Collection of dietary intakes and calculation of Alternative Healthy Eating Index (AHEI-2010) Participants completed semi-quantitative food frequency questionnaires (FFQs) at 8-15 and 32-40 weeks gestation. This FFQ was adapted for pregnant women from the full length Block-98 FFQ (NutritionQuest, Berkeley, CA) and has been validated in pregnant populations (206, 207). The FFQ asks women to report their diet during the previous three months (101). Therefore, dietary intakes at 8-15 and 32-40 weeks gestation reflect maternal diets in early and mid-to-late pregnancy, respectively. Data on dietary intakes at each timepoint were used to calculate early and mid-to-late pregnancy AHEI-2010, which is an 11 component diet quality measure (scored out of 110) based on foods and nutrients predictive of chronic disease risk and mortality (102, 103). The 11 food or nutrient components (six positive and five negative) include: vegetables, fruit, whole grain, nuts/legumes, omega-3 fatty acids (docosahexaenoic acid and eicosapentaenoic acid), polyunsaturated fatty acids, sugar-sweetened beverages/fruit juice, red/processed meat, trans fat, sodium, and alcohol. Based on recommendations from the American College of Obstetrics and Gynecology that pregnant women should abstain from alcohol during pregnancy (386), we removed the alcohol component from the total AHEI-2010 score. Therefore, AHEI-2010 in this study is scored out of 100, and higher AHEI-2010 scores reflect better overall diet quality. The AHEI-2010 has been validated in other pregnancy 247 cohorts, reporting that AHEI-2010 is associated with gestational length and various measures of newborn size at birth (378-380, 387). We used the mean of early and mid- to-late pregnancy AHEI-2010 scores to reflect average pregnancy diet quality as we observed no meaningful differences in associations when considering the two timepoints separately (data not shown). B.4.6. Statistical analysis We used instrumental reading values for non-zero paraben biomarker concentrations below the LOD to avoid bias associated with imputing values for non-detectable concentrations (107). To account for urine dilution, we used the following formula to adjust all urinary paraben biomarker concentrations: Pc = P[(1.016 − 1)/(SG − 1)], where Pc is the specific gravity-adjusted paraben biomarker concentration, P is the measured paraben concentration (ng/mL), 1.016 is the sample median specific gravity, and SG is the specific gravity of each woman’s pooled urine sample (109). We evaluated ethylparaben, methylparaben, and propylparaben as continuous variables that were ln- transformed. Six women had ethylparaben concentrations of zero, so we added a constant (1.0) before ln-transformation to avoid undefined estimates (108). Due to the narrow distribution of urinary butylparaben concentrations that centered around the LOD (Table 22), and because only 44% of women had butylparaben concentrations above the LOD, we excluded butylparaben from further analyses. We presented maternal characteristics by newborn sex and maternal average pregnancy diet quality as n (%) or median (range). Paraben biomarker concentrations from I-KIDS 248 and the National Health and Nutrition Examination Survey (NHANES) cycles 2013-2016 are presented as the median (25th, 75th percentiles) (110, 111) – for comparability, these concentrations were not adjusted for urine dilution. Birth outcomes are reported as the median (25th, 75th percentiles) by newborn sex and maternal diet quality. We tested for statistical differences in maternal characteristics and birth outcomes by newborn sex and diet quality using chi-squared or Fisher’s exact test for categorical variables and Kruskal- Wallis test for continuous variables. We used linear regression models that included a multiplicative interaction between paraben biomarkers and newborn sex to evaluate sex-specific associations of maternal urinary paraben concentrations with continuous gestational length, birth size, and head circumference (objective 1). We were unable to evaluate newborn weight using clinical size-for-gestational-age categories (large, appropriate, or small-for-gestational age) because few newborns were born small-for-gestational age in this cohort. Gestational length, birth weight, birth weight z-scores, body length, weight/length ratio, and head circumference were normally distributed, and we checked regression models for non- constant residual variance to ensure model assumptions were met. We evaluated both unadjusted and adjusted sex-specific associations of paraben biomarkers with birth outcomes, and reported all results stratified by sex regardless of the significance of the interaction P-value (Pint) between parabens and newborn sex. We identified important confounders a priori and using previously published literature that informed a directed acyclic graph (Figure 15) (365). We evaluated correlations (Pearson for continuous and polychoric for categorical variables) between all covariates; however, none of the chosen 249 covariates were strongly correlated with each other (r < 0.4; data not shown). Therefore, adjusted linear regression models accounted for age, race/ethnicity, conception season, pre-pregnancy BMI, parity, depression status in early pregnancy, newborn sex, and average pregnancy diet quality. Pre-pregnancy BMI, EPDS scores, and diet quality were included as continuous variables, while the remaining variables were categorical, with the reference groups indicated in Table 21. Figure 15. Directed acyclic graph for associations of paraben biomarkers with birth outcomes. The green circle with filled triangle represents the exposure (i.e. paraben biomarkers), while the blue circles with filled line represents the outcomes (i.e. gestational length and newborn anthropometrics). The white circles represent latent variables (socioeconomic status, racism, health lifestyle, seasonality, reproductive health, mental health, and pregnancy health), while the red and blue circles represent confounding or moderating variables, respectively, that were included in final adjusted models. Green circle represents variables that are ancestors of the exposure. 250 To evaluate if sex-specific associations of paraben biomarkers with gestational length, birth size, and head circumference differed by maternal diet quality (objective 2), we used linear regression models accounting for the previously mentioned covariates, a three-way interaction, and all relevant two-way interactions between parabens, sex, and AHEI-2010. We dichotomized mean pregnancy AHEI-2010 at the sample median (score of 55.9) to determine if sex-specific associations of parabens with birth outcomes differed in male or female newborns of mothers with poorer (AHEI-2010 < median) or better (AHEI-2010 ≥ median) average diet qualities. We reported all results regardless of the significance of the three-way interaction P-value (Pint) between parabens, sex, and diet quality. We also conducted sensitivity analyses to determine the robustness of associations. First, gestational length could mediate associations of paraben biomarkers with birth weight, body length, weight/length ratio, and head circumference (388), so we also evaluated these relationships by additionally adjusting for gestational age at birth (Table 24). Second, we also conducted sensitivity analyses where we included preterm births (n = 21). Because we observed that associations of paraben biomarkers with birth outcomes did not change after inclusion of preterm births (data not shown), we excluded pre-term births from our analyses. Lastly, for a small number of newborns, we obtained body length and head circumference data from hospital crib cards (n = 28 and 9 measurements, respectively). Ultimately, we included these babies in final statistical models because associations of parabens with birth outcomes did not differ upon their exclusion (data not shown). 251 Because parabens were ln-transformed for all analyses, we back-transformed the resulting β-estimates and 95% confidence intervals (CIs) using the equation [β * ln(2.00)] to represent the change in birth outcome for every two-fold increase in paraben biomarker concentration. All analyses were conducted in SAS version 9.4 (SAS Institute Inc, Cary, NC) using PROC GLM. We focused on patterns of associations rather than statistical significance. We considered associations potentially meaningful if (1) the upper and lower confidence limits did not cross the null or (2) the upper or lower confidence limit did cross the null but the limit was close to zero (117, 118). Based on recommendations from others (166), we did not adjust for multiple comparisons. B.5. RESULTS B.5.1. Sociodemographic and lifestyle characteristics of the I-KIDS sample In this sample, most mothers were older than 30 years (59%), non-Hispanic white (80%), college educated (81%), employed (85%), and did not smoke in the first trimester (87%), while 50% were nulliparous. Around half of mothers were normal weight, whereas the other half had overweight or obesity before pregnancy. Conception season was uniformly distributed over the four seasons, and most women (74%) had a vaginal delivery. Median (range) EPDS score was 4.0 (0.0 – 19.0), while maternal average pregnancy AHEI-2010 score was 51.9 (25.6 – 76.2). Characteristics of the I-KIDS sample by newborn sex and maternal average pregnancy diet quality are presented in Table 21. Out of 458 newborns included in this study, 121 and 112 were females of mothers with worse or better diet quality, respectively, while 108 and 117 were males of mothers with worse or better quality, respectively (Table 21). Maternal age, education, pre-pregnancy BMI, and AHEI- 252 2010 scores significantly differed by newborn sex and maternal diet quality (Table 21). Table 21. Maternal and select delivery/newborn characteristics (n=458). Characteristics Mothers of female newborns Mothers of male newborns Worse diet Better diet Worse diet Better diet quality quality quality quality n = 121 n = 112 n = 108 n = 117 n (%) n (%) n (%) n (%) P2 Age 0.0002 < 30 years (ref) 64 (52.9) 40 (35.7) 53 (49.1) 32 (27.4) ≥ 30 years 57 (47.1) 72 (64.3) 55 (50.9) 85 (72.7) Race/ethnicity 0.66 Non-Hispanic White (ref) 95 (78.5) 94 (83.9) 86 (79.6) 91 (77.8) Others1 26 (21.5) 18 (16.1) 22 (20.4) 26 (22.2) Education <0.0001 Some college or less 30 (24.8) 8 (7.1) 36 (33.3) 13 (11.1) College grad or high 91 (75.2) 104 (92.9) 72 (66.7) 104 (88.9) Employment 0.43 Unemployed 19 (15.7) 11 (9.8) 18 (16.7) 19 (16.2) Employed 102 (84.3) 101 (90.2) 90 (83.3) 98 (83.8) Parity 0.84 Nulliparous (ref) 59 (48.8) 58 (51.8) 51 (47.2) 61 (52.1) Primiparous 37 (30.6) 39 (33.9) 35 (32.4) 38 (32.5) Multiparous 25 (20.6) 16 (14.3) 22 (20.4) 18 (15.4) Smoking in the first trimester 0.06 No 106 (87.6) 98 (87.5) 89 (82.4) 108 (92.3) Yes 7 (5.8) 2 (1.8) 11 (10.2) 4 (3.4) Unknown 8 (6.6) 12 (10.7) 8 (7.4) 5 (4.3) Conception season 0.13 Winter 30 (24.8) 25 (22.3) 37 (34.3) 22 (18.8) Spring (ref) 32 (26.4) 35 (31.3) 29 (26.9) 32 (27.4) Summer 29 (24.0) 22 (19.6) 17 (15.7) 37 (31.6) Fall 30 (24.8) 30 (26.8) 25 (23.1) 26 (22.2) Mode of delivery 0.51 Vaginal delivery 97 (80.8) 78 (70.8) 77 (73.3) 80 (69.0) Scheduled C-section 12 (10.0) 16 (14.6) 15 (14.3) 17 (14.6) Emergency C-section 11 (9.2) 16 (14.6) 13 (12.4) 19 (16.4) Size-for-gestational age 0.04 Small-for-gestational age (SGA) 5 (4.9) 8 (8.3) 0 (0.0) 7 (7.0) Appropriate-for-gestational age (AGA) 85 (82.5) 71 (74.0) 88 (88.0) 77 (77.0) Large-for-gestational age (LGA) 13 (12.6) 17 (17.7) 12 (12.0) 16 (16.0) Median Median Median Median P3 (min, max) (min, max) (min, max) (min, max) Pre-pregnancy BMI (kg/m2) 25.1 (17.5, 52.1) 23.2 (17.1, 41.0) 26.0 (18.2, 44.9) 24.0 (18.2, 48.8) 0.001 EPDS score 4.0 (0.0, 19.0) 4.0 (0.0, 15.0) 4.0 (0.0, 17.0) 4.0 (0.0, 19.0) 0.72 AHEI-2010 score 45.8 (25.6, 51.9) 59.7 (52.0, 75.5) 44.6 (30.7, 51.9) 59.6 (52.0, 76.2) <0.0001 1 Includes Hispanic white, non-Hispanic black, Asians, Native American or Alaska Natives, Native Hawaiians or Pacific Islanders, multiracial or others. 2P-values from chi-squared or Fisher’s Exact test. 3P-values from Kruskal-Wallis test. Percentages may not add up to 100% due to missing. AHEI-2010, Alternative Healthy Eating Index 2010; BMI, body mass index; EPDS, Edinburgh Postnatal Depression Scale. B.5.2. Maternal paraben biomarker concentrations All women had detectable (above the LOD) urinary concentrations of methylparaben, while > 99% of women had detectable urinary levels of at least two paraben biomarkers 253 (Table 22). The order of median paraben biomarker concentrations was as follows: methylparaben > propylparaben > ethylparaben > butylparaben. Moderate-to-strong correlations were observed between methylparaben and propylparaben concentrations (r = 0.70) and between ethylparaben and methylparaben concentrations (r = 0.43; Figure 16). Median I-KIDS maternal urinary paraben biomarker concentrations were lower than those of same-aged women from NHANES during similar time periods (Table 22) (110, 111). Only maternal urinary ethylparaben concentrations significantly differed by newborn sex and maternal diet quality, where mothers who had females and better diet qualities had highest, while those who had males and worse diet qualities had lowest ethylparaben concentrations (Table 23). Table 22. Urinary paraben biomarker concentrations. I-KIDS 2013-2018 NHANES 2013-2016 Detectable Overall sample (n=458) Women 18 – 40 years old (n=742) Biomarker % ≥ LOD Median (25th, 75th percentile) Median (25th, 75th percentile) Butylparaben 43.9 0.1 (10 years 159 (20.7) 280 Table 29 (cont’d). Demographic & lifestyle characteristic n (%)# BMI at 18 years of age (kg/m2)* Underweight (<18.5) 132 (17.2) Normal weight (18.5-24.9) 531 (69.1) Overweight (25-29.9) 65 (8.5) Obese (≥30) 32 (4.2) BMI at 45-54 years of age (kg/m2) Underweight (<18.5) 11 (1.4) Normal weight (18.5-24.9) 291 (37.9) Overweight (25-29.9) 208 (27.1) Obese (≥30) 258 (33.6) Change in BMI from age 18 to 45-54 Remained under/normal weight 294 (38.3) Became overweight/obese 369 (48.1) Became under/normal weight 5 (0.7) Remained overweight/obese 92 (12.0) *Variables selected as potential confounders in logistic and linear regression models based on a directed acyclic graph. BMI, body mass index. #May not add up to 100% due to missing data. C.4.3. Collection of pregnancy history At baseline, women also answered questions about gravidity (number of pregnancies), parity (number of live births), and ages at first birth and last pregnancy. Gravidity and parity were reported as counts and categorized as zero pregnancies/births, one pregnancy/birth, and two-or-more pregnancies/births. Ages at first birth and last pregnancy were reported in years and assessed as continuous variables. C.4.4. Collection of anthropometrics Weight and height were measured without shoes by trained clinic staff and values were rounded to the nearest 0.5 pound and 0.5 inch, respectively. We calculated BMI (kg/m2) using weight (kilograms) and height (meters) and categorized BMI as under-/normal weight (BMI < 25) or overweight/obese (BMI ≥ 25) (450). We also categorized women by their BMI at ages 18 and 45-54 as follows: women who remained normal weight through age 45-54 (under-/normal weight at ages 18 and 45-54), women who became 281 overweight/obese by age 45-54 (under-/normal weight at age 18 but overweight/obese at age 45-54), women who became normal weight by age 45-54 (overweight/obese at age 18 but under-/normal weight at age 45-54), and those who remained overweight/obese through age 45-54 (overweight/obese at ages 18 and 45-54). C.4.5. Assessment of mid-life hormones Hormone analyses are described in detail elsewhere (449). Briefly, fasting morning blood samples were collected at baseline and once per week for four consecutive weeks to assess serum estradiol, testosterone, and progesterone concentrations at all phases of the menstrual cycle (averaged across the four visits). Participants were compensated after each clinic visit for their time and travel to the clinic (449). Hormone concentrations were assessed using commercially available and previously validated ELISA (DRG, Springfield, New Jersey, USA) according to manufacturers’ instructions (440). Mean values of duplicates were used in analyses. Average intra- and inter-assay coefficients of variation were <5% (449, 451). Limits of detection (LOD) for each hormone were as follows: estradiol = 9.71 pg/mL; testosterone = 0.08 ng/mL; progesterone = 0.05 ng/mL (452). Values less than the LOD were assigned the LOD for that hormone. C.4.6. Statistical analysis Three women were excluded from the analysis because of missing information about pregnancy history, baseline weight/height, and/or progesterone concentrations. Another woman was excluded for having extreme testosterone concentrations across all visits. Therefore, 768 pre- and peri-menopausal women were available to assess associations 282 of pregnancy history with mid-life BMI. Binary logistic regression models assessed whether pregnancy history increased or decreased the probability of being overweight/obese compared with being under-/normal weight. To assess whether relations were linear across BMI, we also assessed the associations between pregnancy history and continuous mid-life BMI using linear regression. In linear regression models, mid-life BMI was natural log-transformed to meet normality assumptions. To evaluate whether pregnancy history was associated with a “shift” in BMI status from age 18 to 45-54 (as described above), we aimed to assess whether pregnancy history was associated with changes in BMI across a woman’s reproductive window (from under- /normal weight to overweight/obese or from overweight/obese to under-/normal weight). Only five women who were overweight/obese at age 18 became under/normal weight at age 45-54, so we ultimately only evaluated associations between pregnancy history and the risk of becoming overweight/obese at 45-54 years of age. We did not compare women who were overweight/obese at age 18 and remained overweight/obese at age 45-54 with other groups because these women were obese prior to being pregnant, which would not allow us to ask whether pregnancy history was the cause of their overweight/obesity. Furthermore, eight women did not report their weight at age 18, thus 663 women were available for assessing these associations in final binary logistic regression models. In sensitivity analyses, we examined associations between pregnancy history and BMI gain since age 18 but excluded women whose first birth occurred before age 18 in order to evaluate whether pregnancy history was the potential cause of weight gain/loss since age 18, specifically. 283 Data are presented before and after we controlled for potential confounders selected using previous literature that informed a directed acyclic graph (453), which was similar to the one published by Pirkle et al., 2014 (454). Final covariate-adjusted models included race, maternal smoking status, fertility problems, oral contraceptive use, menopausal status, and BMI at age 18 (except where we assessed BMI gain as our outcome). Associations of gravidity and parity with mid-life BMI or BMI gain were additionally adjusted for age at first birth (only in women who had ever given birth), because age at first birth could impact the number of pregnancies and births a woman could have. Similarly, associations of ages at first birth and last pregnancy with mid-life BMI or BMI gain since age 18 were additionally adjusted for parity (for age at first birth) and gravidity (for age at last pregnancy). To understand whether hormones partially explained (mediated) the proposed relations between pregnancy history and mid-life BMI, we used a system of structural equations controlling for the same confounders mentioned above. Specifically, mediation analyses assessed the total effect (to evaluate the overall relationship between pregnancy history and mid-life overweight/obesity), the natural direct effect (to assess how much of the total relationship is explained by the direct relationship between pregnancy history  mid-life overweight/obesity) and natural indirect effect (to test how much of the total relationship is mediated through hormones: pregnancy history  hormone  mid-life overweight/obesity) (455). All hormone data were natural log-transformed to meet normality assumptions. 284 Because we hypothesized a priori that associations of pregnancy history with mid-life BMI or BMI gain since age 18 could differ between pre- and peri-menopausal women, we also assessed these associations separately by menopausal status. All associations were considered significant at P<0.05 using SAS 9.4 (version 14.3, SAS Institute), including PROC CAUSALMED to evaluate mediation. C.5. RESULTS C.5.1. Characteristics/pregnancy history of MWHS participants At the baseline visit, the majority of women were 45-49 years old (65.2%), pre- menopausal (64.6%), Caucasian/White (65.5%), college graduates (63.8%), employed (79.2%), married or living with a partner (65.0%), and had annual household incomes of ≥$40,000 (74.4%) (Table 29). Physical activity “compared to others” was evenly distributed between “less than others” (33.2%), “as much as others” (31.3%), and “more than others” (34.4%). Most women (65.1%) reported consuming alcohol during the year before the study, whereas 54.7% reported never smoking, and 56.8% reported that their mother did not smoke while pregnant with them. Approximately 81% of women had never sought medical consultation for fertility problems, and 70.9% used oral contraceptives for at least one year (Table 29). Most women had at least two pregnancies (76.8%) or live births (60.2%). The median (range) age at first birth was 27 (12-46) years, whereas the median (range) age at last pregnancy was 33 (14-53) years (Table 30). Although, 69.1% of women reported having a normal weight at age 18, the majority (60.7%) were overweight/obese at 45-54 years. 285 Around half of women did not change BMI categories since age 18, whereas 48.1% of women gained weight from age 18 to 45-54. Median (range) mid-life hormone concentrations were as follows: estradiol: 56.70 (9.71 - 434.52) pg/mL; testosterone: 0.29 (0.08 - 5.40) ng/mL; and progesterone: 1.60 (0.05 - 32.93) ng/mL. Table 30. Pregnancy history of women in the Mid-Life Women’s Health Study (n=768). Pregnancy Characteristic n (%) or Median (range) Gravidity (number of pregnancies) Never pregnant 87 (11.3) 1 pregnancy 91 (11.9) 2 or more pregnancies 590 (76.8) Parity (number of live births) Never pregnant 87 (11.3) No live births 78 (10.2) 1 live birth 139 (18.1) 2 or more live births 462 (60.2) Missing 2 (0.3) Age at first birth (years) 27.0 (12.0 - 46.0) Age at last pregnancy (years) 33.0 (14.0 - 53.0) Data are presented as n (%) or median (range). C.5.2. Associations of pregnancy history with mid-life overweight/obesity We first assessed whether pregnancy history was associated with BMI status at 45-54 years of age, where BMI was assessed in categories (logistic regression analysis) or linearly (linear regression analysis). We found that gravidity was not associated with mid- life overweight/obesity in unadjusted or covariate-adjusted models (Figure 18A). After additionally adjusting for age at first birth, women with two-or-more pregnancies had 62% lower odds of overweight/obesity than women with one pregnancy (OR=0.38; 95%CI: 0.17, 0.83; P=0.02). The linear relationship between gravidity and mid-life BMI was attenuated compared to results from logistic models, where women with two-or-more pregnancies had a non-significant 4.9% lower BMI compared to those with one pregnancy 286 after adjusting for age at last pregnancy (β=-0.05; 95%CI: -0.11, 0.007; P=0.09; Table 31). Figure 18. Associations of gravidity (number of pregnancies) and parity (number of live births) with mid-life overweight/obesity. Binary logistic regression models evaluated associations of A) gravidity and B) parity with the probability of being overweight/obese compared with under-/normal weight in mid-life. Results are expressed as odds ratio (filled diamond) and 95% confidence interval (solid lines) for the unadjusted, adjusted, and additionally adjusted models. Analyses control for race, maternal smoking status, fertility problems, oral contraceptive use, BMI at age 18, and menopausal status, and additionally control for age at first birth. Significant associations are those that do not cross the null (odds ratio=1.0) at #P<0.10, *P<0.05, and **P<0.01. Compared with women who had one live birth, women who never gave birth and those who gave birth two-or-more times had lower odds of being overweight/obese by age 45- 54 years (Figure 18B). In unadjusted models, women who gave birth two-or-more times had 44% lower odds of overweight/obesity than women who only gave birth once (OR=0.56; 95%CI: 0.37, 0.84; P=0.005), and women who had never given birth had 40% lower odds of overweight/obesity than women with one live birth, but this was marginally 287 non-significant (OR=0.60; 95%CI: 0.34, 1.07; P=0.09). After adjusting for potential confounders, women with zero or two-or-more live births had 53% (OR=0.47; 95%CI: 0.23, 0.96; P=0.04) and 42% (OR=0.58; 95%CI: 0.35, 0.95; P=0.03), respectively, lower odds of overweight/obesity than women who only gave birth once. Even after additionally adjusting for age at first birth, women with two-or-more live births had 51% (OR=0.49; 95%CI: 0.29, 0.82; P=0.007) lower odds of overweight/obesity than women with one birth. Linear regression models evaluating these associations differed somewhat from logistic regression analyses (Table 31). In the unadjusted linear model, women with two-or-more live births had 7.1% lower mid-life BMI compared with women with one birth (β=-0.07; 95%CI: -0.12, -0.03; P=0.002), but this was attenuated after adjusting for potential confounders and age at first birth. In linear regression models, mid-life BMI did not differ between women who never gave birth and those who gave birth once. 288 Table 31. Unadjusted and adjusted linear regression analyses evaluating associations between pregnancy history and natural log-transformed mid-life BMI. Model 1a Model 2b Model 3c Modification by β (95% CI) P-value β (95% CI) P-value Menopause Status β (95% CI) P-value P-value# Gravidity No pregnancies -0.02 (-0.09, 0.05) 0.64 0.01 (-0.05, 0.07) 0.70 0.74 1 pregnancy (reference) 0 0 0 2+ pregnancies -0.04 (-0.09, 0.01) 0.14 -0.02 (-0.06, 0.03) 0.47 -0.05 (-0.11, 0.007) 0.09 Parity No live births -0.05 (-0.12, 0.01) 0.12 -0.04 (-0.09, 0.02) 0.19 0.62 1 live birth (reference) 0 0 0 2+ live births -0.07 (-0.12, -0.03) 0.002 -0.03 (-0.06, 0.01) 0.15 -0.03 (-0.07, 0.004) 0.08 Age at first birth -0.009 (-0.01, -0.005) <0.0001 -0.002 (-0.005, 0.001) 0.21 0.33 -0.003 (-0.006, 0.000) 0.09 Age at last pregnancy -0.004 (-0.007, -0.001) 0.01 -0.001 (-0.004, 0.002) 0.46 0.15 -0.001 (-0.003, 0.002) 0.70 β: beta-estimate; CI: confidence interval. aModel 1: unadjusted model; bModel 2: adjusted for race, maternal smoking status, fertility problems, oral contraceptive use, BMI at age 18, and menopause status; cModel 3: model 2 additionally adjusted for age at first birth (gravidity and parity models), parity (age at first birth model), and gravidity (age at last pregnancy model). # Model 2 assessed whether menopause status is an effect modifier of the association between pregnancy history and mid-life BMI. 289 The age at which women gave birth to their first child was associated with mid-life overweight/obesity in all logistic regression models (Figure 19A). In unadjusted models, each year increase in age at first birth was associated with 8% lower odds of mid-life overweight/obesity (OR=0.92; 95%CI: 0.90, 0.95; P<0.0001). After adjusting for potential confounders and additionally adjusting for parity, each year increase in age at first birth was associated with 4% (OR=0.96; 95%CI: 0.92, 1.00; P=0.03) and 6% (OR=0.94; 95%CI: 0.91, 0.98; P=0.005) respectively, lower odds of mid-life overweight/obesity. In unadjusted linear regression models (Table 31), each year increase in age at first birth was associated with a 0.9% decrease in mid-life BMI (β=-0.009; 95%CI: -0.01, -0.005; P<0.0001), which was attenuated after adjusting for confounders. Figure 19. Associations of age at first birth and last pregnancy with mid-life overweight/obesity. Binary logistic regression models evaluated associations of age at A) first birth and B) last pregnancy with the probability of being overweight/obese compared to under/normal weight in mid-life. Results are expressed as odds ratio (filled diamond) and 95% confidence interval (solid lines) for the unadjusted, adjusted, and additionally adjusted models. Analyses control for race, maternal smoking status, fertility problems, oral contraceptive use, BMI at age 18, and menopausal status, and additionally control for parity (for age at first birth) or gravidity (for age at last pregnancy). Significant associations are those that do not cross the null (odds ratio=1.0) at #P<0.10, *P<0.05, and **P<0.01. 290 Age at last pregnancy was associated with mid-life overweight/obesity only in the unadjusted logistic regression model, such that every year increase in age at last pregnancy was associated with 3% lower odds of mid-life overweight/obesity (OR=0.97; 95%CI: 0.94, 1.00; P=0.03; Figure 19B). Similarly, only in the unadjusted linear regression model was age at last pregnancy associated with mid-life BMI, where every one year increase in age at last pregnancy was associated with 0.4% lower mid-life BMI (β=-0.004; 95%CI: -0.007, -0.001; P=0.01; Table 31). Associations between pregnancy history and mid-life overweight/obesity were not different by menopausal status (data not shown). C.5.3. Associations of pregnancy history with BMI gain since age 18 Because we observed associations between pregnancy history and BMI status at age 45- 54, we also wanted to ask whether pregnancy history was associated with BMI change (or more specifically, BMI gain) from age 18 to age 45-54. Our data suggest that parity, but not gravidity, was associated with BMI gain since age 18 (Figure 20). In the unadjusted model, compared with women who gave birth once, those with two-or-more births had 39% lower odds of becoming overweight/obese (OR=0.61; 95%CI: 0.40, 0.93; P=0.02; Figure 20B). This association was marginally non-significant after adjusting for confounders (OR=0.66; 95%CI: 0.41, 1.06; P=0.08). However, after additionally adjusting for age at first birth, women who gave birth two-or-more times had 43% lower odds of becoming overweight/obese in mid-life compared with women who only gave birth once (OR=0.57; 95%CI: 0.34, 0.94; P=0.03). BMI gain since age 18 in women who never gave birth was not significantly different from women who gave birth only once. 291 Figure 20. Associations of gravidity (number of pregnancies) and parity (number of live births) with becoming overweight/obese in mid-life. Binary logistic regression models evaluated associations of A) gravidity and B) parity with the probability of becoming overweight/obese compared with remaining under-/normal weight from age 18 to mid-life. Results are expressed as odds ratio (filled diamond) and 95% confidence interval (solid lines) for the unadjusted, adjusted, and additionally adjusted models. Models controlled for race, maternal smoking status, fertility problems, oral contraceptive use, BMI at age 18, and menopausal status, and additionally controlled for age at first birth. Significant associations are those that do not cross the null (odds ratio=1.0) at #P<0.10, *P<0.05, and **P<0.01. Age at first birth was also associated with BMI gain since age 18 (Figure 21A). In unadjusted models, each year increase in age at first birth was associated with 8% lower odds of becoming overweight/obese in mid-life (OR=0.92; 95%CI: 0.89, 0.95; P<0.0001). After adjusting for confounders and additionally adjusting for parity, each year increase in age at first birth was associated with 3% (OR=0.97, 95%CI: 0.93, 1.00; P=0.06) and 5% (OR=0.95, 95%CI: 0.92, 0.99; P=0.01), respectively, lower odds of becoming overweight/obese in mid-life. Age at last pregnancy was only associated with BMI gain in the unadjusted model, where every year increase in age at last pregnancy was associated with 4% lower odds of becoming overweight/obese in mid-life (OR=0.96; 95%CI: 0.94, 292 0.99; P=0.02; Figure 21B). Associations between age at last pregnancy and BMI gain since age 18 were attenuated after adjusting for confounders and additionally adjusting for gravidity. Associations between pregnancy history and BMI gain since age 18 were not different by menopausal status (data not shown). All associations remained consistent in sensitivity analyses excluding women who gave birth before age 18 (Table 32). Figure 21. Associations of age at first birth and last pregnancy with becoming overweight/obesity in mid-life. Binary logistic regression models evaluated associations of age at A) first birth and B) last pregnancy with the probability of becoming overweight/obese compared with remaining under/normal weight from age 18 to mid-life. Results are expressed as odds ratio (filled diamond) and 95% confidence interval (solid lines) for the unadjusted, adjusted, and additionally adjusted models. Models controlled for race, maternal smoking status, fertility problems, oral contraceptive use, BMI at age 18, and menopausal status, and additionally controlled for parity (for age at first birth) or gravidity (for age at last pregnancy). Significant associations are those that do not cross the null (odds ratio=1.0) at #P<0.10, *P<0.05, and **P<0.01. 293 Table 32. Sensitivity analysis: unadjusted and adjusted logistic regression analyses modeling the probability of becoming overweight/obese (BMI≥25 kg/m2) compared to remaining under/normal weight (BMI<25 kg/m2) at 45-54 years of age (excluding women whose first birth occurred prior to age 18). Model 1a Model 2b Model 3c Modification by OR (95% CI) P-value OR (95% CI) P-value Menopause Status OR (95% CI) P-value P-value# Becoming overweight/obese versus remaining under/normal weight from age 18 to 45-54 years Gravidity No pregnancies -- -- -- -- -- -- -- 1 pregnancy (reference) 1.00 1.00 1.00 2 or more pregnancies 0.55 (0.28, 1.10) 0.09 0.60 (0.28, 1.27) 0.18 0.37 0.56 (0.26, 1.19) 0.13 Parity No live births -- -- -- -- -- -- -- 1 live birth (reference) 1.00 1.00 1.00 2 or more live births 0.60 (0.39, 0.92) 0.02 0.67 (0.41, 1.08) 0.10 0.99 0.59 (0.36, 0.96) 0.04 Age at first birth 0.93 (0.90, 0.96) <0.0001 0.97 (0.93, 1.01) 0.09 0.52 0.96 (0.92, 1.00) 0.03 Age at last pregnancy 0.97 (0.94, 1.00) 0.04 1.01 (0.97, 1.04) 0.72 0.50 1.01 (0.98, 1.05) 0.53 OR: odds ratio; CI: confidence interval. a Model 1: unadjusted model; bModel 2: adjusted for race, maternal smoking status, fertility problems, oral contraceptive use, and menopause status; cModel 3: model 2 additionally adjusted for age at first birth (gravidity and parity model), parity (age at first birth model), and gravidity (age at last pregnancy model). #Model 2 assessed whether menopause status is an effect modifier of the association between pregnancy history and BMI gain since age 18. 294 C.5.4. Associations of pregnancy history and mid-life BMI mediated by mid-life reproductive hormones We previously reported associations of mid-life hormones with mid-life BMI (440). Presently, we observed consistent associations of parity and age at first birth with both mid-life overweight/obesity and BMI gain since age 18. Therefore, we assessed whether mid-life estradiol, testosterone, or progesterone concentrations mediated these associations (Figure 22). While we observed that parity was significantly associated with mid-life overweight/obesity (natural direct effect), there was no mediation of the relationship between parity and mid-life overweight/obesity by estradiol, testosterone, or progesterone (natural indirect effect, Figure 22A). Similarly, age at first birth was also significantly associated with mid-life overweight/obesity (natural direct effect), but there was no mediation of the relationship between age at first birth and mid-life overweight/obesity by estradiol, testosterone, or progesterone (natural indirect effect, Figure 22B). Hormones also did not mediate these associations after concurrently adjusting for parity or age at first birth, and relationships in the mediation analyses were also not modified by menopausal status (data not shown). 295 Figure 22. Associations of parity (number of live births) and age at first birth with mid-life BMI – mediation by mid-life hormones. A system of structural equations assessed the mediating effect of estradiol, testosterone, and progesterone on associations of parity and age at first birth with mid-life BMI. Models were adjusted for race, maternal smoking status, fertility problems, oral contraceptive use, BMI at age 18, and menopausal status. Results are expressed as OR (95%CI) for the total effect, natural direct effect, and natural indirect effect. Green filled arrows represent significant associations (P<0.05), whereas orange unfilled arrows represent non-significant (P>0.05) associations. C.6. DISCUSSION Results from this study suggest that parity and age at first birth are important predictors of overweight/obesity in pre- and peri-menopausal women. Specifically, women who gave birth only once and those who were younger at their first birth had higher odds of being overweight/obese in mid-life, even after controlling for important confounders. Additionally, parity and age at first birth were independently associated with weight gain from age 18 to age 45-54. These associations were consistent between pre- and peri- menopausal women, and associations of parity and age at first birth with mid-life BMI were not explained by mid-life hormones. 296 C.6.1. Parity, but not gravidity, was associated with mid-life overweight/obesity and BMI gain since age 18 Adaptations in maternal carbohydrate and lipid metabolism, especially in mid-to-late pregnancy, lead to gestation-related fat accumulation to central regions (236, 456). This could partially explain why we observed that parity, but not gravidity, was strongly associated with mid-life obesity and adult weight accumulation. Carrying a child to full term is associated with maternal metabolic changes and weight gain that persist after pregnancy. Furthermore, having children is accompanied by unique biological and lifestyle shifts, which put parous women at higher risk of obesity compared with nulliparous women (specifically those who become pregnant but do not give birth to a child). Interestingly, our observation that having one child (being primiparous) puts women at greatest risk of mid-life overweight/obesity is consistent with studies showing that parous women gain the most weight during their first pregnancy/birth compared with subsequent ones (457, 458). This suggests that there is a higher likelihood of drastic and persistent body composition and metabolic changes after having one child. Although the exact causes of this “primiparous paradox” are not clear, studies suggest that multiparous women (compared with those who only have one child) are less likely to experience anxiety and depression after pregnancy (459), and are more motivated to make dietary changes and lose weight after subsequent pregnancies (460), which could put primiparous women at higher risk of obesity. Similar to our observations, a large prospective cohort of U.S. Caucasian/White and African American/Black pre-menopausal 297 women found that women with one live birth (but not two-or-more live births), had higher weight gain and waist-to-hip ratio compared with women with no live births over a five year follow-up (461). Similar findings were observed at the 10 year follow-up of the same cohort, where substantial weight gain occurred in women who had only one birth, whereas higher order births were not associated with excess weight gain (462). These findings are analogous to ours, showing that women who only give birth once have higher odds of mid-life overweight/obesity than women who never give birth or those who give birth two- or-more times. Our results are somewhat inconsistent with several other studies in racially/ethnically diverse populations of only pre- (463, 464), post- (441, 465, 466), or a combination of pre- , peri-, and post-menopausal women (442-447), showing that having more live births is linearly associated with higher BMI or with weight gain since age 18 (444). Given that most women in our study were Caucasian/White or African American/Black and none were post-menopausal, these discrepant findings could be due to racial/ethnic and/or menopausal status differences, which will need to be further investigated. Overall, as discussed above, our findings are supported by previous research, and appear to suggest that women who give birth only once are more likely to experience weight gain during their reproductive years, but substantially more data are needed to understand the social and lifestyle factors that influence this observation. 298 C.6.2. Age at first birth, not age at last pregnancy, was associated with mid-life overweight/obesity and BMI gain since age 18 Pregnancy itself is associated with numerous metabolic changes to support fetal growth, including insulin resistance and fat accumulation (236, 467), and these metabolic changes can persist after pregnancy (468). Additionally, pregnancy and obstetric complications could impact physical activity and mobility well after delivery (469). These deleterious effects would emerge earlier in women who have children at a younger age, which could have major implications for lifelong health. Our finding that women who have their first child at a younger age are at elevated risk for mid-life overweight/obesity is consistent with previous studies in pre-, peri-, and post-menopausal women (441, 463, 470, 471). One study proposed that giving birth at a younger age provides women with more time to have children, leading to weight accumulation between pregnancies (470). The same study also found that among post-menopausal Chinese women, having more reproductive years was associated with higher mid-life BMI and waist circumference (470). In our study, younger age at first birth was associated with higher risk of BMI gain from age 18 to 45-54, and this was independent of parity. Because younger age at first birth has been associated with higher central adiposity (441, 470) and poorer physical performance during mid-life (469), earlier childbirth could be an independent risk factor for mid-life overweight/obesity. Beyond biological factors, earlier childbirth is associated with numerous lifestyle shifts. For example, women who give birth in their teens or 20s might have to stall their education to care for their child (472). Younger mothers might also have lower incomes, leading to 299 poorer diet quality compared to older mothers (473). Although we did not record age at last birth, our findings were consistent with other studies showing no associations between age at last pregnancy and mid-life obesity or weight accumulation (441, 474), suggesting that age at first birth is a stronger predictor of lifetime obesity risk than the age at which a woman stops having children. C.6.3. Hormones did not mediate associations of parity and age at first birth with mid-life overweight/obesity We previously reported that as mid-life BMI increases, mid-life estradiol and progesterone concentrations decrease, whereas testosterone concentrations increase (440). In the current study, we hypothesized that pregnancy-related changes in hormones might persist into mid-life and partially explain observed associations of parity and age at first birth with mid-life overweight/obesity. However, we did not observe such a mediation. The relationship between reproductive hormones and obesity, especially during the menopausal transition, is complex. Studies in pre-, peri-, and post-menopausal women have shown that adipose tissue deposition can be influenced by reproductive hormones (434), but that the reverse can also be true (440, 475). Therefore, additional studies are needed to better understand the mechanisms connecting mid-life obesity and shifts in reproductive hormones. C.6.4. Strengths and limitations This study has several strengths and limitations. Although this cohort is not a true representative sample of U.S. women, we were able to contribute to the growing literature 300 supporting pregnancy history as a predictor of mid-life obesity in U.S. pre- and peri- menopausal women. However, because this study was designed to evaluate predictors of hot flashes in mid-life women, future studies should additionally obtain information about pre-pregnancy weight, gestational weight gain, and lifestyle before and during pregnancies. Although there is potential for bias from pregnancy history self-recall, there is good-to-excellent agreement between pregnancy history recall compared with medical records (476, 477). Additionally, each woman’s mid-life height and weight were measured and recorded by trained clinical staff, which reduced the potential for bias and variability in our outcome. There is also potential for bias with self-reported weight at age 18; one validation study suggest that individuals have a tendency to under-report past body weight; however, the same study suggests that self-recall of past body weight is accurate at the population level (478). Although women in the study were predominately Caucasian/White and African American/Black (and not any other races/ethnicities), this relatively homogeneous population provided us with the power to assess potential hormonal mechanisms driving associations of pregnancy history with mid-life obesity. Lastly, blood samples were not drawn on specific days or phases of the menstrual cycle. Because women were experiencing irregular menses, the lack of standardized blood collection could have introduced variability in hormone measurements. Despite this limitation, four blood samples over four consecutive weeks were collected and averaged to provide better estimates of each woman’s reproductive hormone status. C.7. CONCLUSIONS Findings from this study suggest that having one child or being younger at first childbirth 301 are important and persistent predictors of a woman’s health before and during menopause. We observed that associations between pregnancy history and mid-life BMI are potentially unrelated to mid-life hormone concentrations, suggesting that other unmeasured modifiable/intervenable factors are involved in pregnancy-related mid-life obesity. Therefore, to reduce the prevalence and incidence of mid-life obesity and its associated morbidities in women, primiparous women and those who are younger at their first childbirth, specifically, might benefit from interventions that teach healthy lifestyle habits during their reproductive years. 302 APPENDIX D: URINARY PHTHALATE METABOLITE CONCENTRATIONS AND SERUM HORMONE LEVELS IN PRE- AND PERIMENOPAUSAL WOMEN FROM THE MIDLIFE WOMEN’S HEALTH STUDY This article/appendix has been published in Environment International; Volume 156; Pacyga DC* and Chiang C*, Strakovsky RS, Smith RL, James-Todd T, Williams PL, Hauser R, Meling DD, Li Z, Flaws JA, *These authors contributed equally; Urinary phthalate metabolite concentrations and serum hormone levels in pre- and perimenopausal women from the Midlife Women’s Health Study. Copyright Elsevier (2021); https://doi.org/10.1016/j.envint.2021.106633. D.1. ABSTRACT Phthalate exposure is associated with altered reproductive function, but little is known about associations between phthalate and hormone levels in midlife women. This cross- sectional analysis includes 45–54-year-old pre- and perimenopausal women from Baltimore, MD and its surrounding counties enrolled in the Midlife Women’s Health Study (n = 718). Serum and urine samples were collected from participants once a week for four consecutive weeks to span the menstrual cycle. Serum samples were assayed for estradiol, testosterone, progesterone, sex hormone binding globulin (SHBG), follicle- stimulating hormone (FSH), and anti-Müllerian hormone (AMH), and geometric means were calculated for each hormone across all four weeks. Urine samples were analyzed for nine phthalate metabolites from pools of one-to-four urine samples. Phthalate metabolite concentrations were specific gravity-adjusted and assessed as individual metabolites or as molar sums of metabolites from common parents (di(2-ethylhexyl) 303 phthalate metabolites, ∑DEHP), exposure sources (plastic, ∑Plastics; personal care products, ∑PCP), biological activity (anti-androgenic, ∑AA), and sum of all metabolites (∑Phthalates). We used linear regression models to assess overall associations of phthalate metabolites with hormones, controlling for important demographic, lifestyle, and health factors. We also explored whether associations differed by menopause status, body mass index (BMI), and race/ethnicity. Most participants were non-Hispanic white (67%) or black (29%), college-educated (65%), employed (80%), and had somewhat higher mean urinary phthalate metabolite concentrations than other U.S. women. Overall, the following positive associations were observed between phthalate metabolites and hormones: ∑DEHP (%Δ: 4.9; 95%CI: 0.5, 9.6), ∑Plastics (%Δ: 5.1; 95%CI: 0.3, 10.0), and ∑AA (%Δ: 7.8; 95%CI: 2.3, 13.6) with estradiol; MiBP (%Δ: 6.6; 95%CI: 1.5, 12.1) with testosterone; ∑DEHP (%Δ: 8.3; 95%CI: 1.5, 15.6), ∑Plastics (%Δ: 9.8; 95%CI: 2.4, 17.7), MEP (%Δ: 4.6; 95%CI: 0.1, 9.2), ∑PCP (%Δ: 6.0; 95%CI: 0.2, 12.2), ∑Phthalates (%Δ: 9.0; 95%CI: 2.1, 16.5), and ∑AA (%Δ: 12.9; 95%CI: 4.4, 22.1) with progesterone; and MBP (%Δ: 8.5; 95%CI: 1.2, 16.3) and ∑AA (%Δ: 9.0; 95%CI: 1.3, 17.4) with AMH. Associations of phthalate metabolites with hormones differed by menopause status (strongest in premenopausal women for estradiol, progesterone, and FSH), BMI (strongest in obese women for progesterone), and race/ethnicity (strongest in non- Hispanic white women for estradiol and AMH). We found that phthalate metabolites were positively associated with several hormones in midlife women, and that some demographic and lifestyle characteristics modified these associations. Future longitudinal studies are needed to corroborate these findings in more diverse midlife populations. 304 D.2. KEYWORDS Phthalates; hormones; mid-life women. D.3. INTRODUCTION Phthalates are commonly used to impart strength and flexibility to a variety of plastic products (479, 480). Additionally, low molecular weight phthalates are often used in personal care products to stabilize scents and colors (479, 480). Phthalates are non- covalently bound to the products in which they are used, allowing them to leach from products over time and resulting in human exposure on a daily basis (481, 482). Phthalates used in food and consumer good production can lead to human exposure by ingestion of foods contaminated with phthalates through processing or packaging and dermal absorption through use of phthalate-containing personal care products and clothing (480, 483, 484). Additionally, people undergoing medical procedures are exposed to phthalates directly via medical devices (485, 486). Further, humans are exposed through routes such as inhalation of house dust and air contaminated with phthalates (487). Although phthalate exposure is ubiquitous in humans, exposure levels vary between populations and even sex. Women have higher exposure to phthalates than men, potentially due to their greater use of personal care products compared to men (488, 489). In fact, studies often find that phthalate metabolites are detectable in 99-100% of samples submitted by women (490, 491), making women an especially vulnerable population. Phthalate exposure is of concern because phthalates have been shown to have 305 endocrine disrupting capabilities (492-496). Epidemiological studies have shown that phthalates are associated with altered hormone levels in both men and women (497-500). Although several epidemiological studies have focused on phthalate exposure in a variety of populations, few studies have investigated health outcomes associated with phthalate exposure in mid-life women. Some studies in older women have shown associations between phthalates and health outcomes such as bone mineral density, hot flash experience, and weight change (501-503). However, less is known about the relationship between phthalates and health outcomes during the menopausal transition (i.e. perimenopause) because most studies have thus far investigated women that classify as either pre- or postmenopausal. The transition into the menopausal state is an event known for its hormonal fluctuations and discomforts. This transition begins when the ovaries undergo follicular exhaustion, which results in a shift in the hormonal milieu during the menopausal transition (504). In a cycling woman, the ovary is the primary source of the sex steroid hormones estradiol, progesterone, and testosterone (505). These sex steroid hormones interact with the hypothalamus and pituitary to affect the production of the gonadotropins, follicle stimulating hormone (FSH) and luteinizing hormone (LH), from the pituitary. As the ovary produces fewer sex steroid hormones with age, the negative feedback exerted by the ovarian hormones on the hypothalamus and pituitary is alleviated, leading to an increase in the release of FSH and LH (506). Additionally, in cycling women, anti-Mϋllerian hormone (AMH) is synthesized by cells within small, growing ovarian follicles, leading to high levels of AMH during prime reproductive years (507). Depletion of the ovarian 306 reserve during aging leads to a loss of follicles that produce AMH, and subsequently, AMH levels decline (504). Thus, the hormonal profile of the non-cycling woman (i.e. postmenopausal) can generally be characterized as having lower levels of sex steroid hormones and AMH and higher levels of gonadotropins (429). The primary objective of this study was to address a gap in previous knowledge about the associations between phthalate levels and hormones that fluctuate during the menopause transition. To do so, we investigated the overall associations of common urinary phthalate metabolites with reproductive hormones including estradiol, testosterone, progesterone, sex hormone binding globulin (SHBG), AMH, and FSH in the Midlife Women’s Health Study (MWHS). Because hormone levels may differ in women based on menopause status, midlife body mass index (BMI), and race/ethnicity, the secondary objective of this study was to evaluate differences in associations of phthalate metabolites with reproductive hormones by these characteristics (449, 508-511). D.4. METHODS D.4.1. Midlife Women’s Health Study Cohort The Midlife Women’s Health Study Cohort (MWHS) is a longitudinal population-based study that recruited women from Baltimore, MD (USA) and its surrounding counties between the ages of 45 and 54 in 2006-2015. The full study protocol has been previously published (449). Briefly, women were eligible if they had 3 or more periods within the past 12 months (pre- or perimenopausal), had no history of oophorectomy or hysterectomy, were not on hormone therapy, were not taking botanical supplements to alleviate 307 menopausal symptoms, were not on oral contraceptives, were not pregnant, were not undergoing cancer treatment, and had no history of ovarian, breast, or endometrial cancer. The current study focused on year 1 of MWHS and included a total of 718 women who had information about reproductive hormones, urinary phthalate metabolite concentrations and specific gravity, and covariates (described in the statistical analysis section). All women gave written and informed consent according to procedures that were approved by the University of Illinois Review Board. D.4.2. Demographic and lifestyle characteristics At the baseline visit, participants filled out a detailed questionnaire about their demographic information, as well as lifestyle characteristics such as alcohol consumption, physical activity, and smoking status. Information on racial and ethnic background was obtained by asking women to choose their most representative race/ethnicity from the following options: Caucasian/white, African American/black, Hispanic, Asian, or other. Alcohol consumption was ascertained by asking women whether they consumed 12 alcoholic drinks in the past year (answer: yes, no). Women self-reported their leisure physical activity compared to others, and this was categorized as physically active much more or more than others, as much as others, or less or much less than others. Lifetime smoking status was self-reported as current, former, or never. Women who reported having at least 1 menstrual period within the last 3 months and at least 11 menstrual periods over the last year were considered premenopausal. Women were classified as perimenopausal if they experienced at least one menstrual period over the last year, but not the past 3 months, or if they experienced a menstrual period within the past 3 months, 308 but had experienced 10 or fewer menstrual periods over the last year. At clinic visits, women were asked to list medications (over the counter and prescription) that they were currently taking. Additionally, at the first clinic visit, women had their height and weight measured by clinic staff and values rounded to the nearest 0.5 pound and 0.5 inch. Body mass index (BMI) in kg/m2 was calculated using the National Institutes of Health on-line BMI calculator. D.4.3. Collection and measurement of hormones Women visited the clinic once a week for up to four consecutive weeks for collection of serum samples. Visits to the clinic occurred in the morning to minimize fluctuation in hormones (512, 513). Levels of circulating hormones were measured in serum samples, which were stored at -20 °C prior to measurement. DRG® enzyme-linked immunosorbent assay (ELISA) kits were used to measure levels of estradiol, progesterone, testosterone, and SHBG. Lypocheck® from Bio-Rad Laboratories was used as a control with known values for estradiol, progesterone, testosterone, and SHBG for every assay of these hormones. All samples, controls, and standards were run in duplicates. The limits of detection (LODs) for estradiol, progesterone, testosterone, and SHBG were 9.714 pg/mL, 0.045 ng/mL, 0.083 ng/mL, and 0.77 nmol/L, respectively. The inter- and intra-assay %CVs were all ≤ 10.0, with the exception of estradiol which was ≤ 14.9. Aliquots of serum from the first visit of each patient were submitted to the University of Virginia Center for Research in Reproduction Ligand Assay and Analysis Core for measurement of levels of AMH and FSH. AMH was assayed via ELISA, and FSH was 309 measured via radioimmunoassay (RIA). The LODs for AMH and FSH were 0.2 ng/mL and 0.1 mIU/mL, respectively. The intra- and inter-assay %CVs for AMH were 3.9 and 6.2, respectively. The intra- and inter-assay %CVs for FSH were 3.2% and 4.9%, respectively. D.4.4. Collection and measurement of phthalate metabolites in urine During the same visit in which women donated serum, spot urine samples were also collected. Each woman provided at least one and up to four urine samples (sample number was dependent on the number of clinic visits completed by each woman), which were pooled for each participant to measure phthalate metabolite concentrations. Due to the short half-lives of phthalates in the body and high within person variability of measured concentrations, previous studies have shown that a pooled sample better represents phthalate exposure compared to a single urine sample (162, 262). Urine samples were stored at -20 °C prior to measurement. Phthalates were measured in pooled urine samples via isotope dilution high-performance liquid chromatography negative-ion electrospray ionization-tandem mass spectrometry (HPLC-MS/MS) at the Metabolomics Center of the Roy J. Carver Biotechnology Center at the University of Illinois at Urbana- Champaign. Phthalate metabolites measured included: mono-2-ethylhexyl phthalate (MEHP), mono-(2-ethyl-5-hydroxyhexyl) phthalate (MEHHP), mono-(2-ethyl-5- oxohexyl)phthalate (MEOHP), mono-(2-ethyl-5-carboxypentyl) phthalate (MECPP), mono-(3-carboxypropyl) phthalate (MCPP), mono-benzyl phthalate (MBzP), monoethyl phthalate (MEP), monobutyl phthalate (MBP), and mono-isobutyl phthalate (MiBP). The limits of detection (LOD) for each phthalate metabolite were as follows: MEHP: 0.2 ng/mL; MEHHP: 0.05 ng/mL; MEOHP: 0.1 ng/mL; MECPP: 0.05 ng/mL; MCPP: 0.05 ng/mL; 310 MBzP: 0.05 ng/mL; MEP: 0.1 ng/mL; MBP: 0.05 ng/mL; and MiBP: 0.1 ng/mL. In addition, the intra-assay and inter-assay CVs were below 5%. Further, all standard curves had correlation coefficient values larger than 0.992 and all runs included internal standards. D.4.5. Statistical analysis To evaluate associations of midlife urinary phthalate metabolite concentrations with hormone concentrations, covariates were chosen a priori and based on previous literature that informed a directed acyclic graph (Figure 23). We assessed correlations among all covariates to evaluate potential multicollinearity issues and found that none of the covariates were strongly correlated with each other. Final statistical models evaluating overall associations of urinary phthalate metabolite concentrations with midlife hormones were adjusted for age, race/ethnicity, employment status, education, annual family income, marital status, menopausal status, alcohol consumption, smoking status, physical activity, midlife BMI, and current medication use. Age and income were included as continuous variables, while the other covariates were categorized with reference groups set as shown in Table 33. For our secondary objective, we a priori stratified our analyses as follows: pre- versus perimenopausal women; under-/normal weight (BMI < 25.0 kg/m2), overweight (BMI ≥ 25.0 – 29.9 kg/m2), versus obese (BMI ≥ 30.0 kg/m2) women; and non-Hispanic white versus black/other women. All stratified models included the previously listed covariates. 311 Figure 23. Directed acyclic graph for associations of phthalates with hormones. Green circle represents the exposure (i.e. phthalates), while the blue circle represents the outcome (i.e. hormones). White circles represent latent variables (i.e. socioeconomic status, racism, and healthy lifestyle), while red circles represent confounding variables that were measured in MWHS and included as covariates in final statistical analyses. Urinary phthalate metabolite concentrations and serum hormone concentrations below the LOD were converted to the LOD/√(2). Because estradiol, testosterone, progesterone, and SHBG concentrations were assessed in multiple samples per participant, the geometric means of these hormones were calculated and used in statistical analyses. To account for differences in urine dilution, phthalate metabolite measurements were adjusted for specific gravity (SG) using the following formula: Pc = P[(1.018 − 1)/(SGi − 1)], where Pc is the specific gravity adjusted concentration, P is the measured concentration (ng/mL), 1.018 is the median specific gravity of the overall MWHS population included in this analysis, and SGi is the specific gravity of each woman’s 312 pooled urine sample (109). Specific gravity-adjusted phthalate metabolite concentrations were used to approximate exposure to common phthalate parent compounds. DEHP metabolites (MEHP, MEHHP, MEOHP, MECPP) were molar converted and summed (nmol/mL) to estimate exposure to DEHP (∑DEHP). The concentrations of the other major urinary phthalate metabolites (MCPP, MBzP, MEP, MBP, MiBP) were not molar converted (ng/mL). Additional phthalate sums (nmol/mL) were created to estimate phthalate exposure based on sources of exposure (personal care products, plastics) and potential biological mechanism (anti-androgenic). MEP, MBP, and MiBP were molar summed to estimate exposure to personal care product phthalates (∑PCP), while MCPP, MBzP, MEHP, MEHHP, MEOHP, MECPP were molar summed to estimate exposure to phthalates found in plastics (∑Plastics). Phthalate metabolites that were shown in in vitro and in vivo studies to have anti-androgenic properties (MBzP, MEHP, MEHHP, MEOHP, MECPP, MBP, MiBP) were molar summed to approximate exposure to anti-androgenic phthalates (∑AA) (22, 149, 495). All phthalate metabolites were also molar-converted and summed to estimate total phthalate metabolite concentrations (∑Phthalates). We used linear regression models to assess overall and stratified associations of midlife urinary phthalate concentrations with midlife hormones. We first evaluated overall associations of continuous phthalates with hormones. Both phthalate and hormone concentrations were natural log-transformed to better approximate normality assumptions in these generalized linear regression models. Second, we evaluated dose-response relationships of phthalates with hormones by categorizing urinary phthalate concentrations into quartiles; hormones were transformed as previously described. For 313 our second objective, linear regression models were stratified by menopause status, midlife BMI, and race/ethnicity to evaluate differences in associations between phthalate metabolites and hormones by these factors. All statistical analyses were conducted in SAS 9.4 (version 14.3, SAS Institute) using PROC GLM. In models where phthalates were assessed as continuous measures (objectives 1 and 2), β-estimates and 95% confidence intervals (CIs) were back- transformed using the equation [((2.00)β – 1)*100] to represent a percent change in hormones for each two-fold increase in phthalate concentration. For models where phthalates were categorized in quartiles, β-estimates and 95% CIs were back- transformed using the equation [(eβ – 1)*100] to represent the percent change in hormones among women in quartiles two (Q2), three (Q3), and four (Q4) of urinary phthalate concentrations, compared to the lowest quartile (Q1). We tested for linear trends (Plinear trend) across quartiles by assessing separate linear regression models that treated the ordinal phthalate variables as continuous. For models evaluating stratified associations of phthalates with hormones, we formally tested for effect modification (Pint) in linear regression by including multiplicative interactions between phthalates and menopause status, phthalates and race/ethnicity, and phthalates and BMI. However, we reported all stratified associations regardless of Pint significance. 314 D.5. RESULTS D.5.1. MWHS demographics and lifestyle characteristics At the time of enrollment, all women were between the ages of 45 and 54 years, with 65% of women being 49 years or younger (Table 33). In terms of racial background, 66% were non-Hispanic white and 34% were black or of other racial/ethnic background. The majority of the women were employed (80%), had a college education or higher (65%), and were premenopausal (64%). Most women reported being at least occasional drinkers (66%) and over half had a midlife BMI ≥ 25 kg/m2 (60%). Over half of women were never smokers (54%), 36% were former smokers, and only 10% were current smokers. 315 Table 33. Demographic and lifestyle characteristics of women from MWHS (n=718). Demographic or Lifestyle Characteristic n (%) Age (years) 45 to 49 469 (65.3) 50 to 54 249 (34.7) Race/ethnicity Non-Hispanic white (ref) 477 (66.4) Black/Other1 241 (33.6) Employment status Unemployed 143 (19.9) Employed (ref) 575 (80.1) Education Some college or less 251 (35.0) College graduate or higher (ref) 467 (65.0) Annual family income ($) <20,000 47 (6.5) 20,000 to 39,999 117 (16.3) 40,000 to 99,999 243 (33.8) ≥100,000 311 (43.3) Marital status Single 129 (18.0) Married/Living with Partner (ref) 466 (64.9) Widowed/divorced/separated 123 (17.1) Menopausal status Premenopausal (ref) 461 (64.2) Perimenopausal 257 (35.8) Alcohol consumption status No drinks or <12 drinks over past year 246 (34.3) At least 12 drinks over past year (ref) 472 (65.7) Smoking status Current 72 (10.0) Former 258 (35.9) Never (ref) 388 (54.0) Leisure physical activity compared to others Much more/more (ref) 258 (35.9) As much 223 (31.1) Less/much less 237 (33.0) Body mass index (kg/m2) Under-/normal weight (<25.0) (ref) 288 (40.1) Overweight (≥25.0-29.9) 187 (26.0) Obese (≥30.0) 243 (33.8) Current medication use None 306 (42.6) Any (ref) 412 (57.4) 1Other includes Hispanic, Asian, or other race/ethnicity. MWHS, Midlife Women’s Health Study. 316 D.5.2. MWHS urinary phthalate metabolite concentrations Median (25th, 75th percentile) urinary concentrations of phthalate metabolites are presented in Table 34. Greater than 99% of women had detectable concentrations (≥ LOD) of all urinary phthalate metabolites (data not shown). Median phthalate metabolite concentrations from our study were generally higher than those in a nationally representative sample of 45-54-year-old women from the National Health and Nutrition Examination Survey (NHANES), likely due to different subpopulations of women in the MWHS and NHANES studies and measurements of urinary metabolites by different laboratories. However, it is important to note that the 25th and 75th percentiles overlapped in metabolite levels between the two studies. 317 Table 34. Phthalate metabolite concentrations in MWHS and NHANES. Name Abbreviation MWHS (2006-2015) NHANES (2005-2016) n=718 n=7571 th th Phthalate metabolite Median (25 , 75 percentile) in ng/mL Mono(2-ethylhexyl) phthalate MEHP 4.5 (2.7, 9.4) 1.2 (0.6, 3.1) Mono(2-ethyl-5-hydroxyhexyl) phthalate MEHHP 33.7 (20.3, 58.1) 9.1 (3.4, 22.5) Mono(2-ethyl-5-oxohexyl) phthalate MEOHP 12 (7.3, 22.3) 5.6 (2.1, 13.2) Mono(2-ethyl-5-carboxypentyl) phthalate MECPP 25.8 (15.8, 48.1) 13.4 (5.6, 31.7) Mono(3-carboxypropyl) phthalate MCPP 2.5 (1.3, 5.4) 1.4 (0.6, 3.4) Monobenzyl phthalate MBzP 9.4 (5.4, 16) 4.1 (1.8, 10.4) Monoethyl phthalate MEP 95.4 (47.4, 192) 58.8 (20.0, 179.6) Mono-n-butyl phthalate MBP 19.7 (12.9, 32.6) 11.5 (5.4, 25.3) Mono-isobutyl phthalate MiBP 16.3 (9.8, 26.1) 5.7 (2.6, 13.1) Phthalate molar-converted sum Median (25th, 75th percentile) in nmol/mL Sum of di(2-ethylhexyl) phthalate metabolites ∑DEHP2 0.3 (0.2, 0.5) 0.1 (0.04, 0.2) Sum of all plastic phthalate metabolites ∑Plastics3 0.3 (0.2, 0.6) 0.1 (0.1, 0.3) Sum of all personal care product phthalate metabolites ∑PCP4 0.7 (0.4, 1.3) 0.4 (0.2, 1.2) Sum of all phthalate metabolites ∑Phthalates5 1.1 (0.7, 2) 0.7 (0.3, 1.8) Sum of anti-androgenic phthalate metabolites ∑AA6 0.5 (0.3, 0.8) 0.2 (0.1, 0.5) 1Phthalate metabolite concentrations for 45–54 year-old US women from combined NHANES cycles 2005–06, 2007–08, 2009–10, 2011– 12, 2013–14, and 2015–16. 2∑DEHP: MEHP, MEHHP, MEOHP, MECPP. 3∑PCP: MEP, MBP, and MiBP. 4∑Plastics): MCPP, MBzP, MEHP, MEHHP, MEOHP, MECPP. 5∑Phthalates: all phthalate metabolites. 6∑AA: MBzP, MEHP, MEHHP, MEOHP, MECPP, MBP, MiBP. MWHS, Midlife Women’s Health Study; NHANES, National Health and Nutrition Examination Survey. 318 D.5.3. MWHS plasma hormone concentrations Plasma hormone concentrations from year 1 of the MWHS are presented in Figure 24. Median (range) hormone concentrations were as follows: estradiol, 49.9 pg/mL (6.9 – 349.3); testosterone, 0.3 ng/mL (0.1 – 4.3); progesterone, 0.6 ng/mL (0.05 – 17.7); SHBG, 64.4 nmol/L (9.0 – 264.8); FSH, 11.3 mIU/mL (0.1 – 161.0); and AMH, 0.1 ng/mL (0.1 – 8.3). Figure 24. Hormone concentrations of women from MWHS (n=718). Mid-life A) estradiol, B) testosterone, C) progesterone, D) sex hormone binding globulin (SHBG), E) follicle stimulating hormone (FSH), and F) anti-Mullerian hormone (AMH) concentrations. Results are presented as 1.5 times the interquartile range below and above the 25th and 75th percentiles (lower and upper endpoints of whisker), the 25th and 75th percentiles (lower and upper edges of box), median (line inside box), and mean (diamond). MWHS, Midlife Women’s Health Study. 319 D.5.4. Overall associations between phthalate metabolites with hormones In linear regression models where phthalate metabolites were modeled continuously, select phthalates were positively associated with estradiol, testosterone, progesterone, and AMH, but not with SHBG or FSH (Table 35). Specifically, 2-fold increases in ∑DEHP, ∑Plastics, and ∑AA were associated with 4.9% (95%CI: 0.5, 9.6), 5.1% (95%CI: 0.3, 10.0), and 7.8% (95%CI: 2.3, 13.6) higher estradiol concentrations, respectively. Additionally, each 2-fold increase in MiBP was associated with 6.6% (95%CI: 1.5, 12.1) higher testosterone concentrations, whereas 2-fold increases in MBP and ∑AA were associated with 8.5% (95%CI: 1.2, 16.3) and 9.0% (95%CI: 1.3, 17.4) higher AMH concentrations, respectively. Lastly, 2-fold increases in ∑DEHP, ∑Plastics, MEP, ∑PCP, ∑Phthalates, and ∑AA were associated with 4.6 – 12.9% higher progesterone concentrations. 320 Table 35. Overall linear associations of phthalate metabolites with hormones (n=718). Exposure Estradiol Testosterone Progesterone SHBG FSH AMH % change in hormones (95%CI) MCPP 1.2 (-1.7, 4.3) 0.1 (-3.0, 3.3) 2.4 (-2.1, 7.2) -2.2 (-4.5, 0.2) -1.9 (-6.9, 3.5) 4.2 (-0.1, 8.7) MBzP 1.2 (-2.8, 5.4) -1.3 (-5.4, 3.0) 3.5 (-2.6, 10.0) 1.6 (-1.7, 4.9) -4.4 (-10.9, 2.7) 5.4 (-0.4, 11.6) ∑DEHP 4.9 (0.5, 9.6) 1.4 (-3.1, 6.1) 8.3 (1.5, 15.6) 0.9 (-2.5, 4.5) -3.7 (-10.8, 4.0) 4.1 (-2.1, 10.7) ∑Plastics 5.1 (0.3, 10.0) 1.6 (-3.2, 6.6) 9.8 (2.4, 17.7) 0.9 (-2.8, 4.7) -4.7 (-12.2, 3.4) 5.4 (-1.3, 12.5) MEP -0.2 (-3.1, 2.7) 0.0 (-3.0, 3.0) 4.6 (0.1, 9.2) 0.3 (-2.1, 2.6) 4.2 (-1.0, 9.7) -1.8 (-5.7, 2.3) MBP 2.0 (-3.0, 7.1) 2.9 (-2.3, 8.3) 6.2 (-1.4, 14.4) 1.1 (-2.8, 5.2) -5.2 (-13.1, 3.4) 8.5 (1.2, 16.3) MiBP 4.0 (-0.8, 9.0) 6.6 (1.5, 12.1) 5.3 (-2.0, 13.1) 1.5 (-2.3, 5.5) -2.9 (-10.7, 5.6) 4.9 (-1.9, 12.1) ∑PCP 0.2 (-3.5, 4.1) 0.8 (-3.1, 4.9) 6.0 (0.2, 12.2) 0.9 (-2.1, 4.0) 3.7 (-2.9, 10.9) -0.1 (-5.3, 5.4) ∑Phthalates 2.3 (-2.1, 6.9) 1.1 (-3.4, 5.9) 9.0 (2.1, 16.5) 1.0 (-2.5, 4.6) 2.1 (-5.5, 10.4) 2.0 (-4.2, 8.5) ∑AA 7.8 (2.3, 13.6) 3.5 (-2.0, 9.3) 12.9 (4.4, 22.1) 2.0 (-2.2, 6.4) -6.9 (-15.1, 2.1) 9.0 (1.3, 17.4) Data are presented as the % change in hormones for every 2-fold increase in phthalate metabolite concentration (ng/mL or nmol/mL). Linear regression models adjusted for age, race/ethnicity, employment status, education, income, marital status, alcohol consumption, smoking status, physical activity, medication use, menopausal status, and BMI. CI, confidence interval; AMH, anti-Mullerian hormone; BMI, body mass index; FSH, follicle stimulating hormone; SHBG, sex hormone binding globulin. 321 In analyses where phthalate metabolites were modeled in quartiles, phthalates were associated with all hormones, except for SHBG (Figure 25). Specifically, compared to those in Q1, estradiol concentrations were 18.7% (95%CI: 4.3, 35.0) higher in women at ∑AA Q4 (Plinear trend = 0.07; Figure 25a), whereas progesterone concentrations were 24.4% (95%CI: 1.8, 51.9) and 26.1% (95%CI: 3.8, 53.2) higher, respectively, in women in the highest quartile of ∑Phthalates (Plinear trend = 0.05) and ∑AA (Plinear trend = 0.04; Figure 25c). Compared to those in the lowest quartile, testosterone concentrations were 14.7% (95%CI: 0.0, 31.6) higher in women at MEP Q3 (Plinear trend = 0.81), as well as 18.5% (95%CI: 3.7, 35.5) and 23.0% (95%CI: 7.3, 41.0) higher, respectively, in women at MiBP Q2 and Q4 (Plinear trend = 0.05; Figure 25b). However, testosterone concentrations were 13.1% (95%CI: 0.5, 24.0) lower in women at ∑DEHP Q2 (Plinear trend = 0.28) compared to those in Q1. Compared to those in the lowest quartile, AMH concentrations were 23.1% (95%CI: 2.6, 47.7) and 19.9% (95%CI: 0.1, 43.7) higher in women at MBP Q2 and Q4 (Plinear trend = 0.09), and 20.7% (95%CI: 0.7, 44.6) higher in women at ∑AA Q3 (Plinear trend = 0.03; Figure 25f). Lastly, FSH concentrations were 31.3% higher (95%CI: 4.1, 65.5) higher in MEP Q3 compared to Q1 (Plinear trend = 0.88; Figure 25e). 322 Figure 25. Associations of phthalate metabolites in quartiles with hormones (n=718). Multivariable linear regression models evaluated associations of urinary phthalate concentrations with A) estradiol, B) testosterone, C) progesterone, D) sex hormone binding globulin (SHBG), E) follicle stimulating hormone (FSH), and F) anti- Mullerian hormone (AMH). 323 Figure 25 (cont’d). Data are presented as the difference in hormone concentration (filled circles) and 95% confidence interval (solid lines) comparing phthalate quartiles 2 (Q2), 3 (Q3), and 4 (Q4) to quartile 1 (Q1). Models were adjusted for age, race, employment status, education, income, marital status, alcohol consumption, smoking status, physical activity, medication use, menopausal status, and body mass index. Confidence intervals that do not cross the null are significantly different from quartile 1 at #P<0.10 and *P<0.05. D.5.5. Associations between phthalate metabolites and hormones stratified by menopause status Associations of phthalates with estradiol, FSH, and AMH were only observed in premenopausal women (Table 36), in whom ∑DEHP, ∑Plastics, and ∑AA were positively associated with estradiol concentrations, MBzP, ∑Plastics, and ∑AA were negatively associated with FSH, while ∑AA was positively associated with AMH. Conversely, MiBP was positively associated with testosterone only in perimenopausal women (Table 36). Associations of phthalates with progesterone were observed in both pre- and perimenopausal women, but they differed depending on the phthalate metabolite (Table 36). ∑DEHP, ∑Plastics, and ∑AA were positively associated with progesterone in premenopausal women, whereas ∑Phthalates was positively associated with progesterone in perimenopausal women. 324 Table 36. Associations of phthalate metabolites with hormones stratified by menopause status. Menopause Phthalate Estradiol Pint Testosterone Pint Progesterone Pint SHBG Pint FSH Pint AMH Pint Status % change in hormones for every 2-fold increase in phthalate concentrations Pre 1.2 (-1.8, 4.2) 0.3 (-3.4, 4.2) 3.7 (-1.6, 9.3) -2.5 (-5.2, 0.3) -3.7 (-8.6, 1.5) 3.8 (-2.3, 10.2) MCPP 0.77 0.43 0.35 0.70 0.21 0.50 Peri 2.4 (-4.4, 9.7) -1.7 (-7.4, 4.4) 1.3 (-7.4, 10.9) -1.5 (-6.3, 3.5) 2.3 (-10.0, 16.3) -0.4 (-3.9, 3.2) Pre 1.8 (-2.2, 6.0) 0.4 (-4.5, 5.6) 4.9 (-2.2, 12.5) 1.0 (-2.7, 4.9) -9.0 (-15.0, -2.5) 7.5 (-0.8, 16.5) MBzP 0.96 0.26 0.65 0.59 0.08 0.15 Peri 1.4 (-7.6, 11.3) -4.7 (-12.1, 3.2) 0.0 (-11.4, 12.9) 3.8 (-2.9, 10.9) 7.3 (-9.7, 27.5) 1.9 (-2.8, 6.9) Pre 6.0 (1.5, 10.8) 0.0 (-5.4, 5.7) 9.4 (1.4, 18.2) 1 (-3.1, 5.2) -7.1 (-13.9, 0.3) 6.7 (-2.3, 16.5) ∑DEHP 0.99 0.59 0.83 0.76 0.30 0.08 Peri 2.0 (-7.2, 12.1) 2.9 (-5.1, 11.6) 5.0 (-7.2, 18.8) 1.3 (-5.3, 8.4) 4.2 (-12.5, 24.1) -1.4 (-6.1, 3.5) Pre 5.8 (1.0, 10.8) 0.8 (-4.9, 6.8) 10.8 (2.2, 20.1) 0.7 (-3.5, 5.2) -8.8 (-15.8, -1.2) 8.2 (-1.4, 18.8) ∑Plastics 0.77 0.87 0.89 0.69 0.20 0.06 Peri 3.0 (-7.0, 14.2) 1.8 (-6.8, 11.3) 6.6 (-6.8, 21.9) 1.7 (-5.5, 9.5) 5.5 (-12.8, 27.7) -1.1 (-6.2, 4.2) Pre 1.3 (-1.6, 4.3) -1.2 (-4.8, 2.5) 4.2 (-1.1, 9.7) 0.5 (-2.2, 3.3) 1.7 (-3.4, 7.0) -2.1 (-7.7, 3.9) MEP 0.60 0.13 0.41 0.71 0.37 0.99 Peri -1.4 (-7.4, 5.0) 2.4 (-3.0, 8.1) 7.8 (-0.6, 17) 0.4 (-4.0, 5.0) 7.0 (-4.7, 20.2) -2.5 (-5.6, 0.7) Pre 3.5 (-1.7, 8.9) 1.4 (-4.9, 8.2) 8.4 (-0.9, 18.5) -0.5 (-5.1, 4.4) -7.5 (-15.3, 1.1) 10.2 (-0.5, 22.1) MBP 0.68 0.69 0.45 0.15 0.43 0.32 Peri 1.8 (-8.3, 13.1) 3.8 (-5.2, 13.6) 1.6 (-11.4, 16.6) 5.8 (-1.8, 14.1) -0.1 (-17.8, 21.4) 2.5 (-2.9, 8.2) Pre 3.4 (-1.5, 8.6) 2.7 (-3.4, 9.1) 7.1 (-1.6, 16.5) -0.2 (-4.6, 4.4) -1.2 (-9.2, 7.5) 2.6 (-6.9, 13.1) MiBP 0.51 0.04 0.60 0.14 0.56 0.72 Peri 9.3 (-1.2, 20.9) 14.8 (5.4, 25.2) 4.3 (-8.6, 19.1) 7.3 (-0.2, 15.4) -7.3 (-23.3, 11.9) 3.3 (-1.9, 8.9) Pre 1.8 (-2.3, 6.1) -0.7 (-5.6, 4.6) 5.3 (-2.0, 13.1) 1.0 (-2.8, 4.9) 1.2 (-5.7, 8.7) 1.0 (-7.0, 9.7) ∑PCP 0.70 0.22 0.54 0.67 0.59 0.68 Peri -0.1 (-7.2, 7.5) 3.0 (-3.3, 9.8) 9.3 (-0.7, 20.3) 1.3 (-3.9, 6.8) 5.0 (-8.4, 20.3) -2.0 (-5.6, 1.8) Pre 4.2 (-0.6, 9.2) -0.8 (-6.5, 5.3) 7.9 (-0.7, 17.1) 1.0 (-3.3, 5.5) -1.9 (-9.6, 6.4) 5.4 (-4.1, 15.9) ∑Phthalates 0.64 0.27 0.52 0.72 0.34 0.20 Peri 0.7 (-7.8, 9.9) 3.7 (-3.9, 11.8) 12.5 (0.4, 26.1) 1.1 (-5.1, 7.7) 6.3 (-9.7, 25.1) -2.4 (-6.7, 2.1) Pre 7.7 (2.2, 13.4) 0.7 (-5.7, 7.5) 12.6 (2.8, 23.3) 0.4 (-4.3, 5.4) -9.8 (-17.6, -1.3) 12.3 (1.2, 24.7) ∑AA 0.44 0.29 0.81 0.15 0.38 0.07 Peri 9.2 (-2.9, 22.7) 7.7 (-2.7, 19.2) 12.2 (-3.8, 30.7) 6.8 (-1.8, 16.2) 0.8 (-18.9, 25.5) 0.3 (-5.6, 6.5) Data are presented as % change in hormone concentration for every 2-fold increase in phthalate metabolite concentration (ng/mL or nmol/mL) in pre- and peri-menopausal women from linear regression models adjusted for age, race/ethnicity, employment status, education, income, marital status, alcohol consumption, smoking status, physical activity, medication use, and BMI. In separate models, an interaction between phthalate and menopause status was included to formally test for effect modification by menopause status, and the resulting P-value (Pint) is provided in the table. CI, confidence interval; AMH, anti-Mullerian hormone; BMI, body mass index; FSH, follicle stimulating hormone; SHBG, sex hormone binding globulin. n = 461 and 257 for pre- and peri-menopausal women, respectively. 325 D.5.6. Associations between phthalate metabolites and hormones stratified by BMI Associations of phthalate metabolites with estradiol were only observed in under-/normal weight women (Table 37). Specifically, ∑AA was positively associated with estradiol. Associations of phthalate metabolites with SHBG were only observed in overweight women, in whom MCPP was negatively associated with SHBG (Table 37). Associations of phthalates with progesterone were only observed in obese women, in whom ∑DEHP, ∑Plastics, MEP, ∑PCP, ∑Phthalates, and ∑AA were positively associated with progesterone (Table 37). Associations of phthalates with FSH and AMH were observed in both under-/normal weight and obese women, but they differed depending on the phthalate metabolite (Table 37). Specifically, ∑DEHP and ∑AA were negatively associated with FSH in obese women, but MBzP was negatively associated, while MEP and ∑PCP were positively associated with FSH in under-/normal weight women. Additionally, MBzP was positively associated with AMH in under-/normal weight women, while MBP was positively associated with AMH in obese women. 326 Table 37. Associations of phthalate metabolites with hormones stratified by mid-life BMI. BMI Phthalate Estradiol Pint Testosterone Pint Progesterone Pint SHBG Pint FSH Pint AMH Pint Category % change in hormones for every 2-fold increase in phthalate concentrations Under/Normal 0.8 (-3.5, 5.4) -2.3 (-7.3, 2.9) -1.9 (-8.8, 5.6) -0.3 (-3.6, 3.1) 0.5 (-7.7, 9.4) 1.0 (-6.1, 8.7) Overweight MCPP 1.6 (-4.6, 8.2) 0.90 -0.1 (-6.3, 6.6) 0.60 4.7 (-4.5, 14.7) 0.27 -5.1 (-9.7, -0.3) 0.36 -7.8 (-18.8, 4.7) 0.44 8.4 (-0.2, 17.7) 0.56 Obese 1.1 (-4.4, 6.9) 3.0 (-2.2, 8.5) 5.9 (-2.2, 14.7) -1.8 (-6.4, 3.1) -1.7 (-9.3, 6.6) 5.7 (-1.5, 13.4) Under/Normal 3.0 (-3.0, 9.3) 0.8 (-6.1, 8.1) 7.1 (-2.9, 18.2) 3.5 (-1.0, 8.3) -11.6 (-21.1, -1.0) 10.5 (0.2, 21.9) Overweight MBzP 5.9 (-3.4, 16.1) 0.18 0.7 (-8.4, 10.7) 0.32 4.9 (-8.3, 20.1) 0.52 -0.8 (-7.9, 6.8) 0.63 6.4 (-11.7, 28.2) 0.20 1.9 (-9.8, 15.2) 0.34 Obese -2.7 (-9.4, 4.5) -5.2 (-11.3, 1.2) -3.1 (-12.5, 7.4) 1.0 (-5.1, 7.4) 2.0 (-8.0, 13.0) 3.4 (-5.5, 13.1) Under/Normal 5.7 (-1.3, 13.3) 0.8 (-7.0, 9.2) 4 (-7.1, 16.5) 2.5 (-2.7, 7.9) -2.7 (-14.7, 11.0) 5.9 (-5.4, 18.7) Overweight ∑DEHP 5.1 (-3.3, 14.2) 0.95 4.5 (-4.1, 13.8) 0.52 4.3 (-7.7, 17.8) 0.11 0.2 (-6.3, 7.1) 0.86 2.2 (-13.7, 21.0) 0.24 1.9 (-8.8, 13.8) 0.80 Obese 3.7 (-4.4, 12.6) -1.6 (-8.8, 6.3) 17.9 (5.1, 32.4) 1.1 (-5.8, 8.6) -11.0 (-20.9, 0.0) 5.4 (-5, 16.8) Under/Normal 6.8 (-0.6, 14.7) 1.6 (-6.7, 10.5) 4.9 (-6.9, 18.1) 3.1 (-2.3, 8.9) -5.0 (-17.2, 9.1) 7.5 (-4.6, 21) Overweight ∑Plastics 5.9 (-3.3, 15.9) 0.78 3.7 (-5.5, 13.8) 0.61 6.7 (-6.6, 21.8) 0.15 0.2 (-6.9, 7.7) 0.73 0.7 (-16.2, 21.1) 0.45 3.8 (-7.9, 17.1) 0.82 Obese 2.6 (-6.0, 12.0) -1.0 (-8.7, 7.4) 19.0 (5.2, 34.6) 0.6 (-6.8, 8.5) -9.9 (-20.5, 2.1) 6.2 (-4.9, 18.5) Under/Normal -4.2 (-9.1, 0.9) 1.6 (-4.5, 8) 1.0 (-7.4, 10.2) 0.8 (-3.2, 4.8) 12.7 (2, 24.5) -4.2 (-12.2, 4.5) Overweight MEP 2.3 (-2.7, 7.6) 0.21 2.8 (-2.4, 8.2) 0.81 1.8 (-5.4, 9.6) 0.21 -2.7 (-6.5, 1.3) 0.01 3.4 (-6.6, 14.5) 0.06 -4.2 (-10.3, 2.4) 0.11 Obese 1.2 (-3.7, 6.4) -2.8 (-7.2, 1.9) 9.7 (2.1, 17.7) 2.6 (-1.7, 7.2) -3.3 (-10.1, 3.9) 4.0 (-2.4, 10.8) Under/Normal 5.5 (-2.3, 13.9) 8.3 (-1.0, 18.5) 5.1 (-7.4, 19.3) 2.6 (-3.2, 8.7) -8.7 (-21.1, 5.8) 14.6 (1.1, 30.0) Overweight MBP -4.2 (-14.2, 6.9) 0.50 0.3 (-10.4, 12.4) 0.32 -5.3 (-19.4, 11.3) 0.19 -1.5 (-9.8, 7.6) 0.94 13.8 (-8.9, 42.2) 0.25 -6.1 (-18.8, 8.7) 0.12 Obese 1.0 (-7.0, 9.7) -1.6 (-8.9, 6.3) 12.3 (-0.2, 26.3) 0.8 (-6.2, 8.3) -8.7 (-18.9, 2.8) 12.8 (1.7, 25.1) Under/Normal 2.6 (-5.0, 10.7) 7.6 (-1.5, 17.6) 0.0 (-11.8, 13.4) 0.2 (-5.4, 6.1) 0.0 (-13.5, 15.7) 8.5 (-4.3, 22.9) Overweight MiBP 6.4 (-2.8, 16.5) 0.83 7.3 (-2.2, 17.8) 0.97 9.9 (-3.7, 25.4) 0.36 0.5 (-6.5, 8.1) 0.18 -2.0 (-18.4, 17.9) 0.63 -6.0 (-16.6, 6.0) 0.10 Obese 3.4 (-5.0, 12.5) 3.6 (-4.2, 12.1) 8.4 (-4.0, 22.3) 3.9 (-3.4, 11.8) -5.9 (-16.7, 6.2) 9.1 (-1.9, 21.3) Under/Normal -3.2 (-9.1, 3.0) 3 (-4.3, 10.8) 1.5 (-8.4, 12.6) 1.1 (-3.6, 6.0) 12.6 (0, 26.9) -2.2 (-11.8, 8.5) Overweight ∑PCP 2.9 (-3.9, 10.3) 0.34 4.1 (-3.0, 11.7) 0.84 5.2 (-4.9, 16.4) 0.41 -1.4 (-6.7, 4.3) 0.09 1.5 (-11.8, 16.8) 0.12 -2.5 (-11.1, 6.9) 0.24 Obese 2.3 (-4.3, 9.4) -2.6 (-8.4, 3.6) 10.4 (0.4, 21.3) 3.0 (-2.8, 9.1) -4.7 (-13.4, 4.9) 6.2 (-2.4, 15.5) Under/Normal 0.7 (-6.5, 8.5) 2.5 (-6.1, 11.8) 3.0 (-8.9, 16.4) 2.7 (-2.9, 8.6) 9.9 (-4.6, 26.6) 0.7 (-10.9, 13.8) Overweight ∑Phthalates 4.6 (-3.2, 13.0) 0.78 5.2 (-2.8, 13.9) 0.80 6.5 (-4.9, 19.3) 0.18 -0.4 (-6.5, 6.0) 0.38 1.0 (-13.7, 18.2) 0.14 1.1 (-8.8, 12.1) 0.55 Obese 3.2 (-4.7, 11.9) -2.7 (-9.7, 4.9) 17.2 (4.6, 31.3) 1.6 (-5.3, 8.9) -7.3 (-17.4, 4.1) 6.8 (-3.5, 18.2) Under/Normal 11.0 (2.1, 20.6) 6 (-3.9, 16.9) 8.5 (-5.5, 24.5) 4.4 (-2.0, 11.2) -7.4 (-21.1, 8.7) 12.7 (-1.8, 29.3) Overweight ∑AA 7.0 (-3.4, 18.5) 0.83 5.2 (-5.3, 16.8) 0.42 9.4 (-5.7, 27.0) 0.25 1.9 (-6.1, 10.6) 0.84 2.1 (-17, 25.7) 0.32 1.0 (-11.9, 15.7) 0.33 Obese 5.5 (-3.9, 15.8) -1.0 (-9.2, 8.0) 20.8 (5.9, 37.8) 1.0 (-6.9, 9.5) -13.2 (-24.0, -0.8) 14.3 (1.7, 28.4) Data are presented as the % change in hormone concentration for every 2-fold increase in phthalate metabolite concentration (ng/mL or nmol/mL) in under-/normal weight, overweight, and obese women from linear regression models adjusted for adjusted for age, race/ethnicity, employment status, education, income, marital status, alcohol consumption, smoking status, physical activity, medication use, and menopausal status. In separate models, an interaction between phthalate and BMI was included to formally test for effect modification by mid-life BMI, and the resulting P-value (Pint) is provided in the table. CI, confidence interval; AMH, anti-Mullerian hormone; BMI, body mass index; FSH, follicle stimulating hormone; SHBG, sex hormone binding globulin. n = 288, 187, and 243 for under-/normal weight, overweight, and obese women, respectively. 327 D.5.7. Associations between phthalate metabolites and hormones stratified by race/ethnicity Associations of phthalate metabolites with estradiol, testosterone, and AMH were only observed in non-Hispanic white women (Table 38). Specifically, ∑DEHP, ∑Plastics, and ∑AA were positively associated with estradiol, MiBP was positively associated with testosterone, and MCPP and ∑AA were positively associated with AMH. However, associations of phthalate metabolites with progesterone were observed in both non- Hispanic white and black/other women (Table 38). Specifically, MCPP and ∑Plastics were positively associated with progesterone in black/other women, while ∑Phthalates was positively associated with progesterone in non-Hispanic white women. 328 Table 38. Associations of phthalate metabolites with hormones stratified by race/ethnicity. Race/ Phthalate Estradiol Pint Testosterone Pint Progesterone Pint SHBG Pint FSH Pint AMH Pint Ethnicity % change in hormones for every 2-fold increase in phthalate concentrations Black/Other 4.1 (-1.6, 10.1) -0.1 (-5.1, 5.1) 13.1 (4.1, 22.9) -2.2 (-6.4, 2.1) -5.9 (-13.5, 2.5) 1.0 (-6.3, 9.0) MCPP 0.25 0.89 0.005 0.97 0.36 0.19 Non-Hispanic white 0.1 (-3.4, 3.8) 0.0 (-4.0, 4.1) -2.6 (-7.8, 2.8) -2.0 (-4.8, 1.0) -0.2 (-6.8, 6.9) 6.3 (0.9, 11.9) Black/Other 4.0 (-3.9, 12.6) 4.4 (-2.9, 12.2) 7.0 (-4.9, 20.5) 1.9 (-4.2, 8.2) -1.6 (-12.7, 11.0) 4.8 (-5.7, 16.5) MBzP 0.25 0.23 0.45 0.87 0.96 0.57 Non-Hispanic white -0.9 (-5.5, 3.9) -3.9 (-8.9, 1.3) 1.3 (-5.7, 8.9) 1.0 (-2.8, 5.0) -4.2 (-12.5, 4.8) 5.6 (-1.4, 13.1) Black/Other 3.4 (-5.0, 12.6) -3.5 (-10.7, 4.3) 12.1 (-1.2, 27.3) -1.1 (-7.4, 5.6) -1.2 (-13.1, 12.4) 2.3 (-8.7, 14.7) ∑DEHP 0.49 0.08 0.56 0.52 0.80 0.49 Non-Hispanic white 5.9 (0.7, 11.4) 3.9 (-1.8, 9.9) 6.7 (-1.1, 15.1) 2.2 (-2.0, 6.5) -4.5 (-13.3, 5.1) 5.2 (-2.2, 13.2) Black/Other 4.6 (-4.5, 14.6) -2.5 (-10.2, 6.0) 17.1 (2.3, 34.1) -1.1 (-7.8, 6.1) -2.3 (-14.9, 12.2) 2.2 (-9.6, 15.5) ∑Plastics 0.75 0.13 0.29 0.57 0.90 0.31 Non-Hispanic white 5.5 (0.0, 11.3) 3.6 (-2.5, 10.0) 6.7 (-1.7, 15.7) 2.0 (-2.4, 6.6) -5.2 (-14.4, 5.0) 7.2 (-0.8, 15.9) Black/Other 1.6 (-3.8, 7.4) 0.6 (-4.4, 5.8) 4.2 (-4.1, 13.2) -0.4 (-4.5, 4.0) 1.5 (-6.7, 10.3) 0.7 (-6.5, 8.4) MEP 0.47 0.62 0.91 0.85 0.58 0.45 Non-Hispanic white -1.1 (-4.4, 2.3) -0.4 (-4.1, 3.5) 4.9 (-0.4, 10.4) 0.6 (-2.2, 3.4) 5.2 (-1.4, 12.2) -2.6 (-7.2, 2.4) Black/Other 0.5 (-9.0, 11.0) 0.6 (-8.1, 10.2) 6.1 (-8.6, 23.2) -0.7 (-8.0, 7.2) -5.5 (-18.7, 9.9) 8.3 (-5.2, 23.7) MBP 0.92 0.86 0.83 1.00 0.71 0.96 Non-Hispanic white 1.6 (-4.1, 7.6) 2.3 (-4.1, 9.1) 4.8 (-3.9, 14.4) 1.0 (-3.6, 5.9) -3.7 (-13.7, 7.5) 7.8 (-0.8, 17.1) Black/Other 3.9 (-4.8, 13.4) 0.4 (-7.3, 8.8) 0.1 (-12.3, 14.3) 3.8 (-3.0, 11.1) 4.6 (-8.5, 19.5) 6.5 (-5.4, 19.8) MiBP 0.81 0.19 0.47 0.22 0.29 0.80 Non-Hispanic white 3.2 (-2.5, 9.2) 9.0 (2.3, 16.1) 7.2 (-1.7, 16.9) 0.1 (-4.5, 4.9) -5.7 (-15.4, 5.2) 3.2 (-5.0, 12.1) Black/Other 2.5 (-3.8, 9.3) 0.5 (-5.2, 6.5) 5.0 (-4.6, 15.5) -0.2 (-5.0, 4.8) 0.8 (-8.5, 11.0) 1.9 (-6.5, 11.0) ∑PCP 0.43 0.92 0.76 0.64 0.61 0.68 Non-Hispanic white -1.2 (-5.8, 3.6) 0.8 (-4.5, 6.3) 6.9 (-0.5, 14.9) 1.8 (-2.1, 5.9) 5.3 (-3.8, 15.3) -1.0 (-7.6, 6.1) Black/Other 3.9 (-3.5, 11.9) -0.2 (-6.7, 6.8) 8.7 (-2.7, 21.4) -1.2 (-6.7, 4.5) 0.7 (-10.0, 12.7) 2.6 (-7.1, 13.3) ∑Phthalates 0.74 0.77 0.91 0.38 0.86 0.85 Non-Hispanic white 1.4 (-4.2, 7.2) 1.7 (-4.5, 8.3) 9.5 (0.6, 19.1) 3.0 (-1.6, 7.9) 2.8 (-7.6, 14.4) 2.2 (-5.8, 10.8) Black/Other 8.4 (-2.5, 20.6) -1.1 (-10.3, 9.1) 18.3 (0.9, 38.7) 0.0 (-7.9, 8.6) -3.7 (-18.1, 13.2) 8.9 (-5.6, 25.7) ∑AA 0.93 0.26 0.44 0.79 0.90 0.75 Non-Hispanic white 7.2 (1.0, 13.8) 4.9 (-1.9, 12.1) 10.3 (0.8, 20.8) 2.9 (-2.1, 8.1) -7.2 (-17.3, 4.0) 9.1 (0.0, 19.0) Data are presented as % change in hormone concentration for every 2-fold increase in phthalate metabolite concentrations (ng/mL or nmol/mL) in non-Hispanic white and Black/Other women from linear regression models adjusted for age, employment status, education, income, marital status, alcohol consumption, smoking status, physical activity, medication use, menopausal status, and BMI. In separate models, an interaction between phthalate and race/ethnicity was included to formally test for effect modification by race/ethnicity, and the resulting P-value (Pint) is provided in the table. CI, confidence interval; AMH, anti-Mullerian hormone; BMI, body mass index; FSH, follicle stimulating hormone; SHBG, sex hormone binding globulin. n = 241 and 477 for black/other and white women, respectively. 329 D.6. DISCUSSION In the present study, we found that several phthalate metabolites were positively associated with both sex steroid and protein hormones. This particular trend was unexpected due to previous in vitro and in vivo studies, as well as observational studies suggesting that phthalates inhibit steroidogenesis (489, 495, 499, 514, 515). Previous observational studies evaluated these associations in men and women during their reproductive life, as well as in children, which may account for these discrepancies given that our study population is in midlife. We also found that some associations of phthalate metabolites with hormones differed by menopause status, midlife BMI, and race/ethnicity, which may provide critical information as to which midlife populations may be more susceptible to the endocrine disrupting effects of phthalates. Overall, our results suggest that phthalates may disrupt steroidogenesis through different mechanisms involving more than simple inhibition. D.6.1. Overall associations of phthalate metabolites with hormones We found that phthalates primarily found in plastic food packaging (i.e. ∑DEHP and ∑Plastics) and those shown to have anti-androgenic activity (i.e. ∑AA) share positive, linear associations with estradiol. ∑AA displayed positive relationships with estradiol in women in the third and fourth quartiles as well, further demonstrating the strength of this positive association. These results are consistent with studies conducted in pregnant women and women between the ages of 16 and 45 that have found positive associations between some phthalate metabolites such as MiBP and MBzP (516), as well as MBP (517) with estradiol, all of which are components of the ∑AA measurement used in our 330 study. However, some of our results are inconsistent with some experimental studies showing that phthalate exposure decreases estradiol levels in rodents (495, 518). Our results also differ from a study in Japanese pregnant women and a recent study in pre- and postmenopausal women from NHANES, which showed that DEHP was associated with lower serum estradiol concentrations (500). We also observed associations of phthalates with testosterone and progesterone. While we did not observe an overall linear association between ∑DEHP and testosterone, our findings from quartile analyses showing negative associations between ∑DEHP and testosterone are consistent with experimental studies showing that DEHP has anti- androgenic properties (59, 63-65). However, we also found that MEP (quartile) and MiBP (linear and quartile) were positively associated with testosterone, which is not consistent with most studies (66-68). Although one observational study showed that prenatal MiBP exposure was associated with increased peripubertal testosterone in girls (69), a cross- sectional study using data from NHANES cycles 2013-2016 found that MEP, MiBP, and ∑DEHP were associated with reduced testosterone, and these associations were strongest in 40-60 year old females (22). Our study population acutely targeted women within a narrow age range to capture the menopausal transition, which may also account for discrepancies in our findings. Most notably, we found that ∑DEHP and MEP were positively associated with progesterone, and these were driving the associations observed for ∑Plastics, ∑PCP, ∑Phthalates, and ∑AA with progesterone. However, previous studies in animals and humans found equivocal results regarding these associations as those studies have reported positive and negative associations of 331 phthalates with progesterone (60, 70-72). Overall associations of phthalates with non-steroid hormones (i.e. AMH, FSH, and SHBG) were less frequent. We found that MBP and ∑AA were positively associated with AMH in both linear and quartile analyses. Few studies have investigated associations between phthalates and AMH, but one research group found inverse associations between concentrations of MBP and AMH in follicular fluid (519), but also reported in an earlier study in the same group of women that MBP shared a positive association with serum AMH, similar to what we observed in our population (520). We also observed that MEP was positively associated with FSH in quartile analyses. However, two studies, one in healthy 16-45 year old women and the other in healthy 11-88 year old men found that some phthalate metabolites (but not MEP) were positively associated with FSH (517, 521). Lastly, we observed no associations between phthalates and SHBG. These results are consistent with studies in peripubertal girls and pregnant women (522-524). Overall, our results and those from previous studies further illustrate the complex relationships that phthalates can share with different hormones and that these associations may also differ across populations. However, additional studies, especially in midlife, are needed to corroborate our findings. D.6.2. Differences in associations by menopause status We found that associations of phthalates with estradiol, progesterone, and FSH were strongest in premenopausal women. Namely, ∑AA, ∑Plastics, and ∑DEHP were all positively associated with estradiol and progesterone in premenopausal women. 332 Coinciding with this finding is that ∑AA and ∑Plastics were also negatively associated with FSH in premenopausal women. Inverse relationships between estradiol and FSH are expected due to the negative feedback loop wherein FSH stimulates estradiol production and estradiol in turn suppresses FSH production. Studies have shown that phthalates are capable of modulating steroidogenic enzymes responsible for rate-limiting steps in the steroidogenesis pathway (525-527). Thus, it is possible that these effects may be due to direct phthalate-induced alterations of steroidogenic enzyme and/or activity. We speculate that these effects may be muted or not present in perimenopausal women because the entire hypothalamic-pituitary-gonadal (HPG) axis in perimenopausal women may be less sensitive to phthalate-induced changes or that the ovary itself is less sensitive to phthalate-induced changes due to the transition into menopause. D.6.3. Differences in associations by midlife BMI While we found that associations of phthalates with most hormones differed by midlife BMI, the most consistent associations were observed with progesterone. Most notably, positive associations of ∑DEHP, ∑Plastics, MEP, ∑PCP, ∑Phthalates, and ∑AA were positively associated with progesterone in obese women only. Adipose tissue is metabolically active with the capability to synthesize and metabolize sex steroid hormones (528). Additionally, the link between phthalates and obesity broadens the possibilities for the relationships that may exist between phthalates, adiposity, and hormone levels (529, 530). It is possible that phthalate-induced disruption in one steroidogenic organ (i.e., the ovary or the adipose tissue) can lead to compensatory action by the other. Alternatively, it is possible that subtle actions on both the ovary and 333 the adipose tissue in women with less adipose mass are less detectable than when in overweight and obese women, thus leading to relationships being observed in overweight and obese women only. The complex relationships that are likely to exist between phthalates, adipose tissue, and hormone levels merit further investigation. D.6.4. Differences in associations by race/ethnicity In race/ethnicity stratified analyses, positive associations of phthalate metabolites with estradiol and AMH were consistently strongest in non-Hispanic white women. Comparison to existing literature is difficult due to the lack of studies that investigate the interaction of hormones, race/ethnicity, and phthalates. However, one study that investigated the changes in hormones in different races found that African American women had a more rapid decline in estradiol concentrations during the menopausal transition than non-Hispanic white women (531). Although not a direct comparison, the study partially supports our findings in that we observed many different positive associations between phthalate measures and hormone levels, but we did not observe that black/other women had positive relationships between any phthalate measures and estradiol. However, this finding contrasts somewhat with other studies that have found that African American women have higher estradiol levels than non-Hispanic white women pre- and post-menopause (532, 533). Circulating hormone concentrations can be influenced by body composition and stress, which could also contribute to racial/ethnic differences in measured hormone levels, as well as result in differential impacts of phthalates on hormones in non-Hispanic white versus black women (440, 534). This highlights the need for further investigation into the complex relationships between 334 race/ethnicity, phthalate exposure, and hormones to fully appreciate the vulnerability of certain populations. D.6.5. Strengths and limitations Our study has limitations and strengths. Due to the cross-sectional nature of our analyses, we are unable to make conclusions about temporality of associations between phthalates and hormones. Further, it is possible that some women in our study experienced irregular menstrual cycles, which could impact hormone levels. However, to counterbalance the variability of menstrual cycles and timing during the cycle for collection of samples, we collected four blood samples that represent each week of a woman’s menstrual cycle for hormone assessment and averaged these hormone concentrations for a more stable outcome measure. Further, the majority of the women in our study were either non- Hispanic white or black, leaving other races and ethnicities underrepresented in our study. While we a priori identified and adjusted for important confounders (i.e. sociodemographic characteristics, behavioral factors, and menopausal status), there may be unobserved or unmeasured confounding variables not accounted for in our statistical models, which could bias our observed associations. For example, diet is an important source of phthalate exposure and may also influence circulating hormone concentrations (535- 538). Given that we were unable to control for diet, we may be overestimating associations between phthalates and hormones levels. Selection bias is also possible if participants with higher phthalate levels had certain characteristics that would impact their hormones. If selection bias exists, it could potentially lead to an under- or overestimation of the strength of our observed associations. 335 Major strengths of our study included the use of a pooled sample for assessing urinary phthalate metabolites, which is important given the short half-lives phthalates have in the body. Additionally, we were also powered enough to detect some differences in associations of phthalates metabolites with hormones by menopausal status, BMI, and race/ethnicity, revealing populations that are potentially more susceptible to the endocrine disrupting effects of phthalates. In addition, this was a multi-racial cohort of midlife women and one of the first studies to provide evidence of associations between urinary phthalate metabolites and hormone levels during a time period of rapid hormonal changes for women—midlife. D.7. CONCLUSION Our study found that some phthalates were associated with several critical hormones in midlife women. Specifically, the following positive associations were observed: ∑DEHP, ∑Plastics, and ∑AA with estradiol; MiBP with testosterone; ∑DEHP, ∑Plastics, MEP, ∑PCP, ∑Phthalates, and ∑AA with progesterone; MBP and ∑AA with AMH. Additionally, associations of phthalate metabolites differed by menopausal status, BMI, and race/ethnicity. Specifically, associations of phthalate metabolites with estradiol, progesterone, and FSH were strongest in premenopausal women, with progesterone were strongest in obese women, and with estradiol and AMH were strongest in non- Hispanic white women. Although some of our findings were corroborated by previous studies, many contrasted with the current literature. The variability in strength and direction of association between phthalates and reproductive hormones highlights the need for future studies to investigate a wide range of exposure windows and to elucidate 336 the mechanism(s) through which phthalates may act to disrupt the HPG-axis. 337 APPENDIX E: URINARY PHTHALATE METABOLITE CONCENTRATIONS AND HOT FLASHES IN WOMEN FROM AN URBAN CONVENIENCE SAMPLE OF MIDLIFE WOMEN This article/appendix has been published in Environmental Research; Volume 197; Pacyga DC* and Warner GR*, Strakovsky RS, Smith RL, James-Todd T, Williams PL, Hauser R, Meling DD, Li Z, Flaws JA, *These authors contributed equally; Urinary phthalate metabolite concentrations and hot flashes in women from an urban convenience sample of midlife women. Copyright Elsevier (2021); https://doi.org/10.1016/j.envres.2021.110891. E.1. ABSTRACT Phthalate exposure is associated with altered reproductive function, but little is known about associations of phthalate exposure with risk of hot flashes. To investigate associations of urinary phthalate metabolite levels with four hot flash outcomes in midlife women. A cross-sectional study of the first year of a prospective cohort of midlife women, the Midlife Women's Health Study (2006–2015), a convenience sample from an urban setting. 728 multi-racial/ethnic pre- and perimenopausal women aged 45–54 years. Women completed questionnaires about hot flash experience and provided 1–4 urine samples over four consecutive weeks that were pooled for analysis. Phthalate metabolites were assessed individually and as molar sums representative of common compounds (all phthalates: ƩPhthalates; DEHP: ƩDEHP), exposure sources (plastics: ƩPlastic; personal care products: ƩPCP), and modes of action (anti-androgenic: ƩAA). Covariate-adjusted logistic regression models were used to assess associations of 338 continuous natural log-transformed phthalate metabolite concentrations with hot flash outcomes. Analyses were conducted to explore whether associations differed by menopause status, body mass index (BMI), race/ethnicity, and depressive symptoms. Overall, 45% of women reported a history of hot flashes. Compared to women who never experienced hot flashes, every two-fold increase in ƩPlastic was associated with 18% (OR: 1.18; 95%CI: 0.98, 1.43) and 38% (OR: 1.38; 95%CI: 1.11, 1.70) higher odds of experiencing hot flashes in the past 30 days and experiencing daily/weekly hot flashes, respectively. Some associations of phthalates with certain hot flash outcomes differed by menopause status, BMI, race/ethnicity, and depressive symptoms. This study suggests that phthalates are associated with hot flash experience and may impact hot flash risk in women who are susceptible to experiencing hot flashes. E.2. KEYWORDS Hot flashes; menopause; phthalates; women. E.3. INTRODUCTION Hot flashes are one of the most common symptoms of menopause, but little is known about the risk factors associated with increased risk of hot flashes. Hot flashes are characterized by sudden and transient periods of intense body heat accompanied by flushing, sweating, chills, and anxiety (539). Experiencing hot flashes can impact daily life for symptomatic women for years and results in estimated medical costs of $340 million in the U.S. each year and an additional $27 million in lost work (540, 541). Although the majority of peri- and postmenopausal women experience hot flashes, the dynamics of hot 339 flashes during menopause transition (such as age of onset, duration, intensity, and risk factors) are not well understood (540). Environmental factors (e.g. smoking), physiological factors (e.g. later stage of menopause), and decreasing estrogen levels are known to be associated with increased risk of hot flashes (542). Our own small cross- sectional analysis of a representative sample of 195 midlife women from the Midlife Women’s Health Study (MWHS) indicated that exposure to phthalates may be associated with an increased odds of hot flashes in midlife women (543). Phthalates are a class of synthetic chemicals composed of esters of ortho-phthalic acid with hydrocarbon side chains of varying lengths. Phthalates are used in a wide variety of consumer products, including food contact materials, medical equipment, car interiors, shower curtains, synthetic leather, and children’s toys, as well as fragranced cleaning and personal care products (544). Humans are ubiquitously and unavoidably exposed to phthalates, with 99% of urine samples from the general U.S. population containing phthalate metabolites (255). Due to greater use of personal care products by women compared to men, women typically have higher concentrations of phthalates than men (545, 546). Importantly, studies show that some phthalate metabolites exert toxicity in biological systems including the reproductive system (547-552). In both experimental and observational studies, phthalates have been shown to alter estradiol levels (495, 550, 553, 554). Animal studies have also shown that mixtures of phthalates can impact hormone levels and steroidogenesis in the ovary (495, 526, 555). This is concerning given 340 that midlife women are widely exposed to phthalates through diet and personal care products (495, 518, 548, 551). The molecular mechanisms of action of phthalates to cause hormone disruption are hypothesized to occur through activation of peroxisome proliferator-activated receptors (495, 550). In general, factors that decrease estrogen levels in women are strongly associated with increased incidence of hot flashes (reviewed in (542)). Extensive literature suggests associations between estrogen levels and hot flashes in conditions that cause acute drops in estrogen levels, such as oophorectomy (556). As phthalates are associated with decreased estrogen levels in human and animals, we hypothesized that phthalate exposure may contribute to hot flash experience in women. To our knowledge, no studies have evaluated the impact of phthalate exposure on hot flash risk in women in detail. However, this study expands upon our previous pilot study (543) to include the entire cohort of the MWHS, which contains multiple racial/ethnic groups, and includes additional analyses based on participant characteristics. Therefore, the primary objective of this study was to assess associations of urinary phthalate metabolite levels with hot flash occurrence, frequency, and severity in midlife women enrolled in the first year of the MWHS. Because risk of hot flashes may differ in women based on menopausal status, midlife body mass index (BMI), race/ethnicity, and depression status (542, 557), the secondary objective of this study was to evaluate differences in associations of urinary phthalate metabolite levels with hot flash risk by these characteristics. 341 E.4. MATERIALS AND METHODS E.4.1. Ethical approval All participants gave written informed consent according to procedures approved by the University of Illinois and Johns Hopkins University Institutional Review Boards (file number: 06741). E.4.2. Study population This study was a cross-sectional analysis of data collected in the first year of the MWHS, a prospective cohort study with the overall goal of evaluating risk factors of hot flashes in midlife women. A detailed study protocol of the MWHS has been published previously (449). Briefly, participants were recruited from the city of Baltimore, MD (USA) and surrounding counties from 2006 to 2015. Women were eligible to participate in the study if they were 45-54 years old and pre- or perimenopausal with or without natural hot flashes. Women were excluded if they had a history of hysterectomy or oophorectomy, were currently pregnant, were taking hormone therapy or herbal/other agents for treatment of menopause symptoms, were taking oral contraceptives, were undergoing cancer treatment, or were postmenopausal. Menopausal status was defined using the Stages of Reproductive Aging Workshop + 10 (STRAW+10) criteria (558). Briefly, menopausal status was defined as follows: pre-menopausal women were those who experienced their last menstrual period within the past 3 months and reported 11 or more periods within the past year. Perimenopausal women were those who experienced their last menstrual period within the past year, but not within the past 3 months, or their last menstrual period within the past 3 months and experienced 10 or fewer periods within the 342 past year. Postmenopausal women were those who had not experienced a menstrual period within the past year. A total of 780 women enrolled in the study during year 1. E.4.3. Collection of demographic and lifestyle characteristics At the baseline clinic visit, women completed a detailed questionnaire and had anthropometrics measured by trained staff. Each woman’s weight and height (without shoes) were measured by trained clinic staff, and values were rounded to the nearest 0.5 pound and 0.5 inch, respectively. The baseline questionnaire collected detailed information on demographics, reproductive history, menstrual cycle characteristics, menopausal symptoms, and medical history, as well as physical activity, smoking status, and alcohol use. Women self-reported their age in years and listed the types of prescription medications used. Each woman’s racial/ethnic background was determined using the question “What is your main ethnic/racial background? (Answer, mark only one: (1) Caucasian/White, (2) African American/Black, (3) Hispanic/Latino, (4) Asian, (5) Other)”. Women reported their highest completed grade or year of schooling using the following options: (1) elementary, (2) high school, (3) technical school, (4) college training, or (5) postgraduate. Smoking status was ascertained using the questions “Have you ever smoked cigarettes?” and “Do you still smoke cigarettes?” whereas the question “In the last 12 months have you had at least 12 drinks of any kind of alcoholic beverage?” was used to determine women’s most recent alcohol consumption status. Leisure physical activity was assessed with the question: “In comparison with others my own age, I think my physical activity leisure time is” (choices: much more, more, as much, less, and much less). Women’s depression status was assessed using the Centers for Epidemiologic 343 Studies Depression Scale (CESD) (559), which is a validated depression score that was calculated using 20 questions that asked about how the women were feeling during the past week. E.4.4. Collection and assessment of hot flash outcomes At baseline, a detailed history of hot flashes was collected using a series of validated questions that have been used in the MWHS for over 10 years (449, 557, 560-563). The current study evaluated four hot flash outcomes that were obtained from the following four questions: 1) whether the woman had ever experienced hot flashes, 2) whether she experienced hot flashes in the past 30 days, 3) the usual severity of her hot flashes, and 4) the usual frequency of her hot flashes. Women were first asked “Have you ever had hot flashes?” where hot flashes were defined as “a sudden feeling of heat in the face, neck, or upper part of the chest” with accompanied “reddening or flushing of the skin followed by sweating and chills.” Women who responded “no” to ever experiencing hot flashes were prompted to skip the more detailed hot flash questions and were categorized as “never experiencing hot flashes”. Those who responded “yes” to ever experiencing hot flashes answered whether they experienced hot flashes within the past 30 days (answer: no, yes). Additionally, women who had ever experienced hot flashes were asked (in general) to describe their hot flashes as: mild (sensation of heat without sweating), moderate (sensation of heat with sweating), or severe (sensation of heat with sweating that disrupts usual activity). We categorized severity of hot flashes as either mild or moderate/severe. Similarly, women who had ever experienced hot flashes were asked (in general) to describe their hot flashes as occurring: every hour, every 2–5 h, every 6– 344 11 h, every 12–23 h, 5–6 days per week, 1–2 days per week, 2–3 days per month, 1 day per month, less than 1 day per month, or never. We categorized frequency of hot flashes as either monthly or daily/weekly. E.4.5. Assessment of urinary phthalate metabolites Urinary phthalate metabolite concentrations are the preferred biomarkers of phthalate exposure (564). Humans are exposed to phthalate diesters (i.e. parent compounds), which are rapidly metabolized to monoester metabolites in the body (565). Therefore, epidemiological studies measuring human urinary phthalate metabolites often measure one or more metabolites for each parent phthalate and report as sums of metabolite concentrations based on parent compound, exposure source, or biological activity (29, 566-570). Participants provided spot urine specimens at the initial baseline clinic visit and at visits during the next three consecutive weeks, which were used for urinary phthalate metabolite assessment. Each woman provided samples at 1–4 visits in the 4 week timeframe, which were pooled due to the short half-lives of phthalates in the body and the high daily and weekly intra-variability of measured concentrations (570). Pooled samples were analyzed for the following 9 phthalate metabolites: mono-2-ethylhexyl phthalate (MEHP), mono-(2-ethyl-5-hydroxyhexyl) phthalate (MEHHP), mono-(2-ethyl-5 - carboxypentyl) phthalate (MECPP), mono-(2-ethyl- 5-oxohexyl)phthalate (MEOHP), mono-(3-carboxypropyl) phthalate (MCPP), monobenzyl phthalate (MBzP), monoethyl phthalate (MEP), monobutyl phthalate (MBP), and monoisobutyl phthalate (MiBP). 345 Analyses were performed using isotope dilution high-performance liquid chromatography negative-ion electrospray ionization-tandem mass spectrometry (HPLC–MS/MS) at the Metabolomics Lab of the Roy J. Carver Biotechnology Center, University of Illinois at Urbana-Champaign, using methods adapted from the Center for Disease Control and Prevention (571) and are described in the supporting information (572). E.4.6. Statistical analysis Out of 780 women enrolled during year 1 of the MWHS, 10 were missing information about hot flashes, an additional 10 were missing information about urinary phthalate metabolite concentrations and/or specific gravity, and an additional 32 women were missing information about covariates (which are described below). Therefore, the current study included a total of 728 women with information about baseline covariates, urinary phthalate metabolite concentrations, and hot flashes. Covariates for associations of midlife urinary phthalate metabolite concentrations with risk of hot flashes were chosen a priori and using previous literature that informed a directed acyclic graph (542, 557). To reduce potential for multicollinearity issues, we assessed correlations among all selected covariates (with none being strongly correlated). Therefore, final statistical models evaluating overall associations of urinary phthalate metabolite concentrations with four hot flash outcomes (objective 1) were adjusted for age, race, education, current drinking status, smoking status, medication use, menopause status, BMI, and CESD score. Age and CESD score were included as continuous variables, whereas the other variables were categorized with reference group set as shown in Table 39. Our second objective was to assess differences in associations of urinary phthalate metabolite concentrations 346 with 4 hot flash outcomes by menopausal status, BMI, race/ethnicity, and CESD score (objective 2) – factors that may influence the experience of hot flashes in midlife women (542). In addition to including the previously listed covariates in all stratified models, we a priori stratified our analyses as follows: pre- versus perimenopausal women, under- /normal weight (BMI<25kg/m2) versus overweight/obese (BMI≥25kg/m2) women, non- Hispanic white versus Black/other women, and women with fewer (CESD<16) versus more (CESD≥16) depressive symptoms (559). Nine urinary phthalate metabolites were assessed from pooled urine samples. Urinary phthalate metabolite concentrations below the level of detection (LOD) were converted to the LOD/√(2). To account for urine dilution, we used the following formula to adjust all urinary phthalate metabolite concentrations: Pc = P[(1.018 − 1)/(SGi − 1)], where Pc is the specific gravity adjusted phthalate metabolite concentration, P is the measured phthalate metabolite concentration (ng/mL), 1.018 is the median specific gravity of MWHS population included in this analysis, and SGi is the specific gravity of each woman’s pooled urine sample (109). Specific gravity-adjusted urinary phthalate metabolite concentrations were used to approximate women’s midlife exposure to phthalate parent compounds. Exposure to DEHP was approximated as the molar converted sum of 4 urinary metabolites using the following equation: ƩDEHP = (MEHP/278) + (MEHHP/294) + (MEOHP/292) + (MECPP/308). Exposure to DOP, BzBP, DEP, DBP, and DiBP was approximated directly using the non-molar converted concentrations of their major urinary metabolites MCPP, MBzP, MEP, MBP, and MiBP, respectively. Additional phthalate sums were created based on primary sources of phthalate exposure, and sum of plasticizer 347 (ƩPlastic) and personal care product (ƩPCP) phthalate metabolites were estimated as follows: ƩPlastic = (MEHHP/294) + (MEHP/278) + (MEOHP/292) + (MECPP/308) + (MCPP/252) + (MBzP/256) and ƩPCP = (MEP/194) + (MBP/222) + (MiBP/222). Previous experimental and some epidemiological studies suggest that certain phthalate metabolites have anti-androgenic activity in the body (149, 495, 573). Therefore, the sum of anti-androgenic phthalate metabolites (ƩAA) was calculated as (MEHP/278) + (MEHHP/294) + (MEOHP/292) + (MECPP/308) + (MBzP/256) + (MBP/222) + (MiBP/222). Lastly, all 9 urinary phthalate metabolites were molar converted and summed to approximate total midlife phthalate exposure (ƩPhthalates). We used logistic regression models to evaluate overall and stratified associations of midlife urinary phthalate metabolite concentrations with the following 4 hot flash outcomes: 1) ever experiencing hot flashes, 2) experiencing hot flashes in the past 30 days, 3) experiencing daily/weekly or monthly hot flashes, and 4) experiencing moderate/severe or mild hot flashes (Table 40). All statistical analyses were conducted in SAS 9.4 (version 14.3, SAS Institute) using PROC LOGISTIC. Specifically, binary logistic regression models assessed overall and stratified associations of continuous midlife urinary phthalate levels with the odds of ever experiencing and experiencing hot flashes in the past 30 days compared to never experiencing hot flashes. Women who did experience hot flashes at some point, but not in the past 30 days (n=87) were excluded from binary logistic regression models comparing women who experienced hot flashes in the past 30 days to those who never experienced hot flashes. Multinomial logistic regression models assessed overall and stratified associations of continuous midlife 348 urinary phthalate levels with the odds of experiencing daily/weekly or monthly hot flashes and of experiencing moderate/severe or mild hot flashes compared to never experiencing hot flashes. Because all phthalate individual metabolites and molar sums were right- skewed, phthalates were natural log-transformed in all logistic regression models. All odds ratios (ORs) and 95% confidence intervals (CIs) were back transformed using the equation [eln(OR)*ln(2.00)] and data in Table 42, Figures 26-29, and the Supplementary Tables (included in (572)) are presented as the OR of experiencing these hot flash outcomes with every two-fold increase in phthalate metabolite or molar sum concentration with a priori alpha level of P < 0.05 (572). For logistic regression models evaluating stratified associations of phthalates with hot flashes, we provided formal test for effect modification in the Supplementary Tables (included in (572)), but reported on all relevant associations regardless of the interaction P-value. E.5. RESULTS E.5.1. Baseline MWHS population characteristics and hot flash prevalence Baseline characteristics of 728 women in the MWHS are reported in Table 39. Overall, 64% of women were premenopausal, whereas 36% were perimenopausal. Most women were white (67%), whereas 33% were Black or of another race/ethnicity. Most women were employed (80%), college educated (65%), married or cohabiting (65%), and had an annual family income ≥$40,000 (75%). The prevalence of baseline healthy lifestyle characteristics in this study population were as follows: 35% did not regularly consume alcohol within the past year, 55% never smoked, 67% reported leisure time physical activity as more/much more than others, 43% were not taking any medications, and 80% 349 did not meet criteria for depression based on the CESD. Over one-third of the women were under- or normal weight (40%) and over half were overweight or obese (60%). Table 39. Demographic and lifestyle characteristics of 45-54 year-old women from the Midlife Women’s Health Study (n=728). Demographic or Lifestyle Characteristic n (%)2 Age (years)1 45 to 49 477 (65.5) 50 to 54 251 (34.5) Race1 Non-Hispanic White (ref) 484 (66.5) Black 215 (29.5) Other 29 (4.0) Employment status Unemployed 146 (20.1) Employed 582 (79.9) Education1 Some college or less 255 (35.0) College graduate or higher (ref) 473 (65.0) Annual family income ($) <20,000 45 (6.2) 20,000 to 39,999 113 (15.5) 40,000 to 99,999 241 (33.1) ≥100,000 308 (42.3) Marital status Single 133 (18.3) Married/Living with Partner 476 (65.4) Widowed/divorced/separated 118 (16.2) Menopausal status1 Premenopausal (ref) 468 (64.3) Perimenopausal 260 (35.7) Alcohol consumption status (>1 drink/month on average)1 No 252 (34.6) Yes (ref) 476 (65.4) Smoking status1 Current 69 (9.5) Former 256 (35.2) Never (ref) 403 (55.4) Leisure physical activity compared to others Much more/more 257 (35.3) As much 230 (31.6) Less/much less 235 (32.3) Body mass index (kg/m2)1 <25 290 (39.8) ≥25 (ref) 438 (60.2) 350 Table 39 (cont’d). Demographic or Lifestyle Characteristic n (%)2 Current medication use1 No 311 (42.7) Yes (ref) 417 (57.3) CES depression score1 Fewer depressive symptoms (<16) 581 (79.8) More depressive symptoms (≥16) 147 (20.2) 1Variables included in logistic regression models. 2Percentages may not add up to 100% due to missing values. The self-reported baseline prevalence of hot flashes in the MWHS is presented in Table 40. Out of 728 total women in the current study, all women provided information about ever experiencing hot flashes, 722 had information about experiencing hot flashes within the past 30 days, 697 had information about hot flash frequency, and 720 had information about hot flash severity. Approximately 55% of women never experienced hot flashes, whereas 45% had experienced hot flashes and 32% had experienced hot flashes in the past 30 days. Overall, 22% had daily/weekly hot flashes and 19% had monthly hot flashes, whereas 29% had moderate/severe hot flashes and 15% had mild hot flashes. 351 Table 40. Prevalence of hot flashes self-reported by women from the Midlife Women’s Health Study (n=728). Hot Flashes n (%) History of hot flashes No 399 (54.8) Yes 329 (45.2) Hot flashes during past 30 days Never had hot flashes 399 (54.8) Had hot flashes and experienced in past 30 days 236 (32.4) Had hot flashes but did not experience in past 30 days 87 (12.0) Missing 6 (0.8) Frequency of hot flashes Never had hot flashes 399 (54.8) Monthly hot flashes 139 (19.1) Daily/weekly hot flashes 159 (21.8) Missing 31 (4.3) Severity of hot flashes Never had hot flashes 399 (54.8) Mild hot flashes 108 (14.8) Moderate/severe hot flashes 213 (29.3) Missing 8 (1.1) E.5.2. Baseline urinary phthalate metabolite concentrations Table 41 presents median (25th, 75th percentiles) concentrations of individual urinary phthalate metabolites and phthalate molar sums during the first year of the MWHS. Concentrations of most urinary phthalate metabolites measured in the MWHS were ≥LOD in 100% of women, except for MEP, for which 99.7% of women had concentrations ≥LOD. Median urinary phthalate metabolite and molar sum concentrations in the MWHS were compared to those measured from a nationally representative sample of 45-54-year-old U.S. women from the 2005-2016 National Health and Nutrition Examination Survey (NHANES). Median metabolite levels were slightly higher in the MWHS than NHANES with overlapping 25–75th percentiles. Metabolite levels were similar to recently reported results from the Study of Environment, Lifestyle, and Fibroids (SELF) from non-Hispanic Black women (23–35 years old) (163). SELF also reports slightly higher DEHP metabolites than the corresponding NHANES samples (163). 352 Table 41. Concentrations of individual urinary phthalate metabolites and molar sums from 45-54-year-old women from the Midlife Women’s Health Study and the National Health and Nutrition Examination Survey. MWHS NHANES Name Abbreviation (2006-2015) (2005-2016) n=728 n=7571 th th Phthalate Median (25 , 75 percentiles) in metabolite ng/mL Mono(2-ethylhexyl) phthalate MEHP 4.5 (2.7, 9.3) 1.2 (0.6, 3.1) Mono(2-ethyl-5-hydroxyhexyl) phthalate MEHHP 33.5 (20.5, 58.7) 9.1 (3.5, 22.6) Mono(2-ethyl-5-oxohexyl) phthalate MEOHP 12.0 (7.3, 22.4) 5.6 (2.1, 13.3) Mono(2-ethyl-5-carboxypentyl) phthalate MECPP 25.9 (15.9, 48.0) 13.4 (5.6, 31.7) Mono(3-carboxypropyl) phthalate MCPP 2.5 (1.3, 5.4) 1.5 (0.6, 3.4) Monobenzyl phthalate MBzP 9.4 (5.4, 16.1) 4.2 (1.8, 10.4) Monoethyl phthalate MEP 97.3 (48.2, 192.0)2 58.9 (20.0, 179.8) Mono-n-butyl phthalate MBP 19.8 (13.0, 32.8) 11.6 (5.4, 25.3) Mono-isobutyl phthalate MiBP 16.4 (10.0, 26.1) 5.8 (2.6, 13.2) Phthalate molar- Median (25th, 75th percentiles) in converted sum3 nmol/mL Di(2-ethylhexyl) phthalate DEHP 0.3 (0.2, 0.5) 0.1 (0.04, 0.2) Sum of all phthalate metabolites ƩPhthalates 1.2 (0.7, 2.1) 0.7 (0.3, 1.8) Sum of all personal care product phthalate metabolites ƩPCP 0.7 (0.4, 1.3) 0.4 (0.2, 1.2) Sum of all plastic phthalate metabolites ƩPlastic 0.3 (0.2, 0.6) 0.1 (0.1, 0.3) Sum of anti-androgenic phthalate metabolites ƩAA 0.5 (0.3, 0.8) 0.2 (0.1, 0.5) 1Weighted phthalate metabolite concentrations for 45-54-year-old US women from combined NHANES survey years 2005-06, 2007-08, 2009-10, 2011-12, 2013-14, and 2015-16. 2Two samples (0.3%) were < LOD for MEP. E.5.3. Overall associations of phthalates with hot flashes Midlife urinary phthalate metabolite concentrations were not associated with ever experiencing hot flashes or experiencing moderate/severe or mild hot flashes compared to never experiencing hot flashes (Table 42). However, phthalates were associated with experiencing hot flashes in the past 30 days. Specifically, women had 19% higher odds of the experiencing hot flashes in the past 30 days with every two-fold increase in MEHHP (OR: 1.19; 95%CI: 1.00, 1.43). Additionally, phthalates were associated with experiencing daily/weekly, but not monthly hot flashes. Specifically, women had 23–38% higher odds of experiencing daily/weekly hot flashes with every two-fold increase in the DEHP metabolites MEHHP (OR: 1.34, 95%CI: 1.09, 1.65), MEOHP (OR: 1.23, 95%CI: 1.03, 1.48), MECPP (OR: 1.28, 95%CI: 1.06, 1.54) and the summary measures ƩDEHP (OR: 353 1.32, 95%CI: 1.08, 1.61), ƩPlastic (OR: 1.38; 95%CI: 1.11, 1.71), ƩPhthalates (OR: 1.26; 95%CI: 1.03, 1.54), and ƩAA (OR: 1.36; 95%CI: 1.07, 1.73). 354 Table 42. Overall associations of urinary phthalate concentrations with hot flashes. Outcomes Experiencing hot Ever experiencing Experiencing daily/weekly or Experiencing moderate/severe or flashes in the past hot flashes1 monthly hot flashes2 mild hot flashes2 30 days1 n 728 635 697 720 Never experiencing Never experiencing Reference Never experiencing hot flashes Never experiencing hot flashes hot flashes hot flashes Daily/weekly Monthly Moderate/severe Mild 3 Exposure OR (95% CI) OR (95% CI) OR (95% CI) OR (95% CI) OR (95% CI) OR (95% CI) MEP 1.09 (0.98, 1.22) 1.05 (0.93, 1.18) 1.1 (0.96, 1.27) 1.1 (0.96, 1.25) 1.06 (0.94, 1.2) 1.12 (0.97, 1.3) MBP 1.01 (0.85, 1.21) 1.1 (0.89, 1.35) 1.17 (0.93, 1.47) 0.92 (0.74, 1.16) 1 (0.81, 1.23) 0.98 (0.76, 1.25) MiBP 0.98 (0.82, 1.16) 1.03 (0.85, 1.26) 1.03 (0.82, 1.3) 0.98 (0.79, 1.21) 1 (0.82, 1.22) 0.91 (0.72, 1.16) ΣPCP 1.08 (0.94, 1.23) 1.04 (0.89, 1.22) 1.12 (0.94, 1.33) 1.07 (0.91, 1.27) 1.03 (0.88, 1.2) 1.12 (0.93, 1.34) MCPP 0.92 (0.83, 1.03) 0.94 (0.82, 1.06) 1.02 (0.88, 1.18) 0.87 (0.76, 1) 0.92 (0.81, 1.05) 0.92 (0.79, 1.07) MBzP 0.95 (0.82, 1.1) 0.93 (0.79, 1.11) 1.08 (0.89, 1.3) 0.94 (0.79, 1.12) 0.95 (0.81, 1.13) 0.94 (0.77, 1.15) MEHP 1.05 (0.92, 1.19) 1.07 (0.92, 1.24) 1.18 (0.99, 1.4) 0.99 (0.85, 1.17) 1.05 (0.9, 1.22) 1.05 (0.88, 1.25) MEHHP 1.14 (0.97, 1.33) 1.19 (1, 1.43) 1.34 (1.09, 1.65) 1.09 (0.9, 1.32) 1.17 (0.98, 1.39) 1.1 (0.88, 1.36) MEOHP 1.07 (0.93, 1.23) 1.11 (0.95, 1.3) 1.23 (1.03, 1.48) 1.04 (0.87, 1.23) 1.07 (0.91, 1.25) 1.08 (0.9, 1.31) MECPP 1.12 (0.97, 1.3) 1.17 (0.99, 1.38) 1.28 (1.06, 1.54) 1.09 (0.92, 1.3) 1.13 (0.96, 1.33) 1.12 (0.93, 1.37) ΣDEHP 1.13 (0.96, 1.31) 1.18 (0.99, 1.42) 1.32 (1.08, 1.61) 1.08 (0.89, 1.3) 1.14 (0.96, 1.37) 1.1 (0.89, 1.36) ΣPlastic 1.12 (0.95, 1.32) 1.18 (0.98, 1.43) 1.38 (1.11, 1.71) 1.05 (0.85, 1.29) 1.14 (0.95, 1.37) 1.1 (0.88, 1.38) ΣPhthalates 1.13 (0.96, 1.32) 1.12 (0.93, 1.35) 1.26 (1.03, 1.54) 1.09 (0.89, 1.33) 1.07 (0.89, 1.29) 1.19 (0.97, 1.47) ΣAA 1.08 (0.9, 1.31) 1.16 (0.93, 1.44) 1.36 (1.07, 1.73) 1 (0.79, 1.27) 1.09 (0.88, 1.35) 1.07 (0.83, 1.39) 1Binary logistic regression models evaluated associations of every 2-fold increase in urinary phthalate concentrations with the odds of ever experiencing and experiencing hot flashes in the last 30 days compared to never having hot flashes. 2Multinomial logistic regression models evaluated associations of every 2-fold increase in urinary phthalate concentrations with the odds of experiencing daily/weekly or monthly hot flashes and experiencing moderate/severe hot flashes compared to never experiencing hot flashes. All logistic regression models were adjusted for age, race/ethnicity, education, alcohol consumption, smoking status, medication use, menopausal status, body mass index, and CESD score. CI, confidence interval; OR, odds ratio. 3ΣDEHP = (MEHP/278) + (MEHHP/294) + (MEOHP/292) + (MECPP/308); ΣPhthalates = (MEP/194) + (MBP/222) + (MiBP/222) + (MEHP/278) + (MEHHP/294) + (MEOHP/292) + (MECPP/308) + (MBzP/256) + (MCPP/252); ΣPCP = (MEP/194) + (MBP/222) + (MiBP/222); ΣPlastic = (MEHHP/294) + (MEHP/278) + (MEOHP/292) + (MECPP/308) + (MCPP/252) + (MBzP/256); ΣAA = (MEHP/278) + (MEHHP/294) + (MEOHP/292) + (MECPP/308) + (MBzP/256) + (MBP/222) + (MiBP/222) 355 E.5.4. Associations of phthalates with hot flashes stratified by menopause status Associations of urinary phthalate metabolite concentrations with ever experiencing hot flashes, experiencing hot flashes in the past 30 days, or experiencing daily/weekly hot flashes did not significantly differ by menopause status (Figure 26A, B, Supplementary Tables (572)). Figure 26. Associations of urinary phthalate concentrations with hot flashes stratified by menopause status. Binary logistic regression models evaluated associations of urinary phthalate concentrations with the odds experiencing hot flashes in the last 30 days compared to never having hot flashes (n=635), while multinomial logistic regression models evaluated associations of urinary phthalate concentrations with the odds of experiencing daily/weekly or monthly hot flashes compared to never experiencing hot flashes (n=697). All models were stratified by menopause status and were adjusted for age, race/ethnicity, education, alcohol consumption, smoking status, medication use, body mass index, and CESD score. Data are presented as odds ratio (filled circles) and 95% confidence interval (solid lines) for every two-fold increasing in urinary phthalate concentrations. Confidence intervals that do not cross the null are significant at #P<0.10 and *P<0.05. 356 E.5.5. Associations of phthalates with hot flashes stratified by midlife BMI Associations of urinary phthalate metabolite concentrations with ever experiencing hot flashes, experiencing hot flashes in the past 30 days, and experiencing moderate/severe or mild hot flashes were not different by midlife BMI (Figure 27A, Supplementary Tables (572)). However, some associations of phthalates with experiencing daily/weekly (but not monthly) hot flashes were only observed in under-/normal weight women (Figure 27B, Supplementary Tables (572)), who had 57–107% higher odds of experiencing daily/weekly hot flashes with every two-fold increase in MBP (OR: 1.57; 95%CI: 1.04, 2.38), MBzP (OR: 1.63; 95%CI: 1.18, 2.25), ƩDEHP (OR: 1.77; 95%CI: 1.24, 2.52), ƩPlastic (OR: 2.07; 95% CI: 1.40, 3.05), and ƩAA (OR: 2.05; 95%CI: 1.31, 3.19). 357 Figure 27. Associations of urinary phthalate concentrations with hot flashes stratified by midlife BMI. Binary logistic regression models evaluated associations of urinary phthalate concentrations with the odds of experiencing hot flashes in the last 30 days compared to never having hot flashes (n=635), while multinomial logistic regression models evaluated associations of urinary phthalate concentrations with the odds of experiencing daily/weekly or monthly hot flashes compared to never experiencing hot flashes (n=697). All models were stratified by midlife BMI and were adjusted for age, race/ethnicity, education, alcohol consumption, smoking status, medication use, menopause status, and CESD score. Data are presented as odds ratio (filled circles) and 95% confidence interval (solid lines) for every two-fold increasing in urinary phthalate concentrations. Confidence intervals that do not cross the null are significant at #P<0.10 and *P<0.05. BMI, body mass index. E.5.6. Associations of phthalates with hot flashes stratified by race/ethnicity Some associations of urinary phthalate metabolite concentrations with ever experiencing hot flashes, experiencing hot flashes in the past 30 days, experiencing daily/weekly or monthly hot flashes, and experiencing mild (but not moderate/severe) hot flashes also differed by race/ethnicity (Figure 28A, B, Supplementary Tables (572)). Non-Hispanic 358 white women had 19-27% higher odds of ever experiencing hot flashes with every two- fold increase in MEP (OR: 1.19; 95%CI: 1.03, 1.38) and ƩPhthalates (OR: 1.27; 95%CI: 1.01, 1.58) (572). Additionally, non-Hispanic white women had 37-45% higher odds of experiencing daily/weekly hot flashes with every two-fold increase in ƩDEHP (OR: 1.37; 95%CI: 1.06, 1.76), ƩPlastic (OR: 1.45; 95%CI: 1.11, 1.90), ƩPhthalates (OR: 1.40; 95%CI: 1.06, 1.85), and ƩAA (OR: 1.40; 95%CI: 1.04, 1.89), as well as higher odds of experiencing monthly hot flashes with every two-fold increase in MEP (OR: 1.23; 95%CI: 1.03, 1.47) (Figure 28B, Supplementary Tables (572)). Conversely, Black/other women had 20% (OR: 0.80; 95%CI: 0.66, 0.98) and 28% (OR = 0.72, 95% CI 0.55, 0.95) lower odds of ever experiencing or experiencing mild hot flashes, respectively, with every two- fold increase in MCPP, but had 51% higher odds of experiencing hot flashes in the past 30 days with every two-fold increase in MEHHP (OR: 1.51; 95%CI: 1.08, 2.10) (572). 359 Figure 28. Associations of urinary phthalate concentrations with hot flashes stratified by race/ethnicity. Binary logistic regression models evaluated associations of urinary phthalate concentrations with the odds of experiencing hot flashes in the last 30 days compared to never having hot flashes (n=635), while multinomial logistic regression models evaluated associations of urinary phthalate concentrations with the odds of experiencing daily/weekly or monthly hot flashes and compared to never experiencing hot flashes (n=697). All models were stratified by race/ethnicity and were adjusted for age, education, alcohol consumption, smoking status, medication use, menopause status, body mass index, and CESD score. Data are presented as odds ratio (filled circles) and 95% confidence interval (solid lines) for every two-fold increasing in urinary phthalate concentrations. Confidence intervals that do not cross the null are significant at #P<0.10 and *P<0.05. E.5.7. Associations of phthalates with hot flashes stratified by CESD score Some associations of urinary phthalate metabolite concentrations with ever experiencing hot flashes, experiencing hot flashes in the past 30 days, experiencing daily/weekly or monthly hot flashes, and experiencing mild (but not moderate/severe) hot flashes differed by CESD score (Figure 29A, B, Supplementary Tables (572)). For example, women 360 with fewer depressive symptoms (CESD<16) had 24% higher odds of experiencing hot flashes in the past 30 days with every two-fold increase in ƩDEHP (OR: 1.24; 95%CI: 1.02, 1.49) (572), and had 38-45% higher odds of experiencing daily/weekly hot flashes with every two-fold increase in ƩDEHP (OR: 1.38; 95%CI: 1.09, 1.73), ƩPlastic (OR: 1.45; 95%CI: 1.14, 1.86), and ƩAA (OR: 1.42; 95%CI 1.07, 1.88) (Figure 29B, Supplementary Tables (572)). Conversely, women with more depressive symptoms (CESD≥16) had 42- 71% higher odds of experiencing hot flashes in the past 30 days with every two-fold increase in MEP (OR: 1.42; 95%CI: 1.02, 1.97), ƩPCP (OR: 1.50; 95%CI: 1.03, 2.19), and ƩPhthalates (OR: 1.71; 95%CI: 1.04, 2.79) (Figure 29A, Supplementary Tables (572)). Additionally, these women had 47-93% and 60-78% higher odds of experiencing daily/weekly or monthly hot flashes, respectively, for every two-fold increase in MEP (daily/weekly OR: 1.47; 95%CI: 1.03, 2.11; monthly OR: 1.60; 95%CI: 1.10, 2.33), ƩPCP (daily/weekly OR: 1.59; 95%CI: 1.06, 2.41; monthly OR: 1.60; 95%CI: 1.05, 2.46), and ƩPhthalates (daily/weekly OR: 1.93; 95%CI: 1.13, 3.28; monthly OR: 1.78; 95%CI: 1.04, 3.06), but also had 43% lower odds of experiencing monthly hot flashes with every two- fold increase in MBzP (OR: 0.57; 95%CI: 0.35, 0.94) (Figure 29B, Supplementary Tables (572)). 361 Figure 29. Associations of urinary phthalate concentrations with hot flashes stratified by CESD score. Binary logistic regression models evaluated associations of urinary phthalate concentrations with the odds of experiencing hot flashes in the last 30 days compared to never having hot flashes (n=635), while multinomial logistic regression models evaluated associations of urinary phthalate concentrations with the odds of experiencing daily/weekly or monthly hot flashes compared to never experiencing hot flashes (n=697). All models were stratified by CESD score and were adjusted for age, race/ethnicity, education, alcohol consumption, smoking status, medication use, menopause status, and body mass index. Data are presented as odds ratio (filled circles) and 95% confidence interval (solid lines) for every two-fold increasing in urinary phthalate concentrations. Confidence intervals that do not cross the null are significant at #P<0.10 and *P<0.05. E.6. DISCUSSION In this cross-sectional analysis of year 1 data from the MWHS, a prospective cohort of pre- and perimenopausal women from Baltimore and its surrounding counties, urinary phthalate metabolite concentrations were associated with experiencing recent and experiencing more frequent hot flashes, but not associated with ever experiencing hot 362 flashes or hot flash severity. Generally, we found that phthalate metabolites of parent compounds found in plastics were associated with increased risk of experiencing hot flashes in the past 30 days and experiencing daily/weekly hot flashes. Some associations of phthalates with certain hot flash outcomes were different by menopause status, midlife BMI, race/ethnicity, and depressive symptoms. Most notably, we found that associations of personal care product phthalates with most hot flash outcomes were strongest in women with more depressive symptoms. Interestingly, MCPP (and to a lesser extent MBzP) was associated with lower risk of hot flashes, especially in perimenopausal and Black/other women. Overall, these results are consistent with our previous findings that phthalate metabolites are associated with increased frequency of hot flashes in midlife women (543). In our pilot study of a representative sample of 195 women from the MWHS cohort, ƩPCP was associated with higher odds of ever experiencing hot flashes, experiencing hot flashes in the past 30 days, and experiencing daily/weekly hot flashes (543). In the current study, we found that associations of ƩPCP with risk of hot flashes only emerged when associations were stratified by CESD score. In the pilot study, ƩAA and ƩDEHP were not associated with hot flash frequency. However, in the current study, we found that ƩDEHP and ƩPlastic were consistently associated with experiencing hot flashes in the past 30 days and hot flash frequency, whereas ƩAA was only associated with hot flash frequency. The pilot study did not evaluate associations of individual phthalate metabolites, ƩPhthalates, or ƩPlastic with hot flash experience. 363 To our knowledge, no other observational studies have reported on the associations between phthalates and hot flashes in midlife women. However, phthalates are associated with disorders of the female reproductive system, including premature reproductive senescence and altered hormone levels, and these disorders may contribute to the risk of hot flashes (542, 557, 561, 574, 575). Studies in mice have shown premature reproductive senescence following adult exposure to DEHP and diisononyl phthalate (DiNP) (576, 577). Observational studies have identified associations between urinary phthalate metabolite levels and earlier onset of menopause (578) and premature ovarian failure (517, 579). Women who undergo earlier menopause report more frequent and severe hot flashes than premenopausal women (575, 580). In addition, higher urinary DEHP metabolite levels were associated with decreased levels of the hormones inhibin B and anti-Müllerian hormone, markers of ovarian reserve in women (581). Hot flashes and declining estradiol levels occur simultaneously during menopause and other physiological transitions such as the postpartum period (542). Clinical trials have shown that estrogen therapy can alleviate hot flashes, suggesting a causal association, although the mechanism remains unknown (510, 582, 583). Numerous observational studies have also identified associations of altered estradiol and progesterone levels with risk of hot flashes (542, 557, 561, 574, 584). Using the MWHS cohort, the close relationship between hot flashes and estradiol was recently modeled using a Bayesian network (510). Animal studies have demonstrated altered estradiol and progesterone levels following phthalate exposure (495, 518, 548, 555, 585), and future studies from the MWHS will investigate the association between urinary phthalate metabolites and 364 hormone levels in midlife women. Other potential mechanisms through which phthalates could be linked to hot flashes include direct disruption of hypothalamic or thyroid function (542, 586, 587). However, as the etiology of hot flashes is not well understood, it is difficult to speculate on causal links between phthalates and hot flashes. When we evaluated stratified associations of phthalates with these hot flashes outcomes by menopause status, BMI, and race/ethnicity, the strongest associations were observed in perimenopausal women, non-Hispanic white women, and under-/normal weight women. These sub-group associations were most consistent between DEHP metabolites and experiencing daily/weekly hot flashes. Black women generally have higher phthalate exposure than white women (588, 589), which may be due to the phthalate content in personal care products used by Black women (590). In the MWHS, Black women had higher urinary PCP phthalate concentrations than non-Hispanic white women (data not shown). However, we did not identify any consistent associations of ƩPCP with risk of hot flashes in Black/other women despite this being a high-risk exposure group, which may indicate that phthalates in personal care products do not contribute to hot flash experience in non-white women. Previously, we have identified high BMI as a risk factor for experiencing perimenopausal hot flashes (574). However, we found stronger associations between phthalates and risk of hot flashes in women with lower BMI. Interestingly, MBzP was associated with increased risk of hot flashes in normal weight women and trending towards decreased risk of ever experiencing hot flashes in overweight women. These differences suggest that women with higher body weight may be less susceptible to endocrine disruption by phthalates. One possible explanation is 365 that estrogen levels are already decreased in perimenopausal women with high BMI (574). Previous studies have identified bidirectional associations of depression with risk of hot flashes, and suggest that they may be linked through sleep disruption (591). In addition, phthalates may be associated with depression in adults. MECPP, MBP, MiBP, and MBzP were associated with depression in adults from NHANES (592). In elderly populations (ages 59-93), DEHP metabolites, MCPP, and MBP have also been associated with depression (593, 594). When associations of phthalates with risk of hot flashes were stratified by CESD scores in the MWHS, we observed different associations of phthalate metabolites, with risk of hot flashes in women experiencing fewer and more depressive symptoms. DEHP metabolites and phthalates found in plastics were consistently associated with hot flashes in women experiencing fewer depressive symptoms, whereas MEP and phthalates associated with personal care products were associated with hot flashes in women with more depressive symptoms. The different observations in women in more vs. fewer depressive symptoms suggests that depression symptoms may be related to hormonal changes or that the physiology of depression may play a role, possibly hormonal, in phthalate mechanism of action. Additional studies are needed to investigate the role of depression in environmental chemical action. In addition, the interaction of each depression group with a different phthalate category may be evidence that the high molecular weight phthalates found in plastics act through different mechanisms than the low molecular weight phthalates found in personal care products. 366 Across multiple hot flashes measures, MCPP and to a lesser extent MBzP were associated with decreased risk of experiencing hot flashes. MCPP is a downstream oxidized metabolite and may be produced from multiple phthalates including MBP, BzBP, and phthalates with long n-hydrocarbon side chains (595). One hypothesis for the negative association between MCPP and hot flashes is that the presence of highly oxidized metabolites is a marker of the overall efficacy of metabolism and detoxification. Greater capacity to metabolize and excrete phthalates (and other environmental chemicals) could reduce their effects on multiple sensitive endpoints. MBzP, containing a benzyl group on its side chain, has a unique structure for a phthalate that is more similar to steroid hormones. As a result, MBzP may act through different mechanisms than other phthalate metabolites. This study has several strengths. Of note, four urine samples taken over consecutive weeks at similar times of day were pooled for phthalate measurement. Within-subject pooling has been shown to decrease exposure misclassification for phthalates and improves the credibility of estimated exposure based on urine concentrations (162). Other strengths of this study include the large size of the population and the detailed information collected from each participant on hot flash experience using validated questionnaires that are accepted by the National Institute of Health to assess hot flashes (596). This study also has several limitations. The MWHS cohort is composed of primarily white (67%) and Black (30%) midlife women. Few women of other races/ethnicities were enrolled in the study; therefore the results of this study are most applicable to non- 367 Hispanic white and Black women. Due to the cross-sectional analyses performed here, some outcomes may have occurred before phthalate exposure, obscuring temporality. The measure of “ever experiencing hot flashes” is the most imprecise and most likely to differ in temporal ordering with respect to phthalate exposure. Therefore, prospective studies are needed to confirm whether phthalates influence risk of experiencing hot flashes. We were unable to adjust for co-pollutants or diet, although we adjusted for a number of relevant and important confounders, including age, race, education, current drinking status, smoking status, medication use, menopause status, BMI, and CESD score. However, other environmental chemical exposures may be correlated with phthalates and hot flashes. Diet quality may be important given that we observed strong associations between plasticizing phthalates and hot flashes risk and exposure to plasticizing phthalates occurs primarily through diet. In addition, we did not evaluate non- linear associations. E.7. CONCLUSION In midlife women from the MWHS, some urinary phthalate metabolites were associated with higher risk of recently experiencing and experiencing frequent hot flashes, but not of ever experiencing or experiencing severe hot flashes. We observed that urinary phthalate metabolites of plasticizer parent compounds were associated with higher odds of experiencing hot flashes in the past 30 days and experiencing daily or weekly hot flashes. Additionally, we found that some associations of urinary phthalate metabolites with hot flashes were different by menopause status, midlife BMI, race/ethnicity, and CESD score. Although this is one of the first studies to assess the relationship between phthalate 368 exposure and risk of hot flashes, these results are consistent with previous studies showing that phthalates can interfere with normal female reproductive function (568). Our results suggest specific relationships between phthalates from common exposure sources and the evaluated hot flash outcomes. Future studies should investigate the mechanisms through which phthalates may be acting to facilitate the development of interventions to alleviate hot flashes in midlife women. 369 APPENDIX F: MIDLIFE URINARY PHTHALATE METABOLITE CONCENTRATIONS AND PRIOR UTERINE FIBROID DIAGNOSIS This article/appendix has been published in International Journal of Environmental Research and Public Health; Volume 19 Issue 5; Pacyga DC, Ryva BA, Nowak RA, Bulun SE, Yin P, Li Z, Flaws JA, Strakovsky RS. Midlife Urinary Phthalate Metabolite Concentrations and Prior Uterine Fibroid Diagnosis. Copyright MDPI (2022); https://doi.org/10.3390/ijerph19052741. F.1. ABSTRACT Fibroid etiology is poorly understood but is likely hormonally mediated. Therefore, we evaluated associations between midlife phthalates (hormone-altering chemicals) and prior fibroid diagnosis, and considered differences by weight gain status. Women (ages: 45–54; n = 754) self-reported past fibroid diagnosis. We pooled 1–4 urines collected after fibroid diagnosis over the consecutive weeks to analyze nine phthalate metabolites and calculate relevant molar sums (e.g., di(2-ethylhexyl) phthalate, ΣDEHP; anti-androgenic phthalates, ΣAA; all metabolites, ΣPhthalates). Using Poisson regression, we evaluated associations between phthalate biomarkers and the risk of having fibroid diagnosis. We explored if associations differed by weight gain from age 18 to 45–54 or in women diagnosed with fibroids within 5 years of phthalate assessment. Our major finding was that women had a 13% (RR: 1.13; 95%CI: 1.02, 1.26) and 16% (RR: 1.16; 95% CI: 1.03, 1.31) greater risk of prior fibroid diagnosis for each two-fold increase in ΣDEHP or ΣAA, respectively. These associations were strongest in women who became overweight/obese from age 18 to 45–54 and in those diagnosed <5 years before phthalate 370 assessment. Based on these results, prospective studies should corroborate our findings related to associations between phthalates and fibroids, and may consider evaluating the role that weight gain may play in these associations. F.2. KEYWORDS Phthalates; endocrine disruptors; fibroids; leiomyoma; midlife; women. F.3. INTRODUCTION By midlife, most women will have uterine leiomyomata, commonly known as fibroids, which are non-cancerous tumors of uterine smooth muscle cells associated with adverse health outcomes, including abnormal uterine bleeding and miscarriage (597-599). Fibroids almost exclusively occur in reproductive-aged, pre- and perimenopausal women, with incidence increasing with age until women are post-menopausal (598, 600, 601). However, the exact prevalence of fibroids is difficult to determine because of the spectrum of clinical presentation. Many women with fibroids have a benign presentation, where fibroids are incidentally detected during imaging, whereas up to half of women with fibroids experience symptoms that are serious enough to impact quality of life, including excessive menstrual bleeding and pelvic pain (602, 603). With very severe symptoms, fibroids are the number one reason for hysterectomy in the United States (602). Given the detrimental impacts of fibroids on women’s quality of life, substantially more data are needed to identify modifiable risk factors that contribute to the development of fibroids. 371 Pre- and perimenopausal women are widely exposed to endocrine disrupting chemicals, including phthalates, which are found in many consumer products. For example, di(2- ethylhexyl) phthalate (DEHP) is a plasticizer used during food processing and in food contact materials, whereas diethyl phthalate (DEP) is used as a fragrance stabilizer in personal care products and cosmetics (240). Given that phthalates are metabolized and excreted within 24-48 hours of exposure, phthalate exposure is best approximated by measuring urinary concentrations of phthalate metabolites (262). Although most individuals in the U.S. general population are exposed to phthalates (604), women have higher measured urinary levels of phthalate metabolites than men, likely due to the use of personal care products and cosmetics (546). In women, studies have demonstrated that phthalates are associated with greater risk of hormone-mediated health outcomes, such as endometriosis (605), hot flashes (250), metabolic syndrome (606), breast cancer (607), and uterine fibroids (608). In cell and experimental animal models, phthalates can alter circulating sex-steroid hormone concentrations by binding to hormone receptors, including peroxisome proliferator-activated receptor alpha receptors and estrogen receptors (alpha and beta) (20, 21, 189, 609, 610). Similarly, in pregnant populations, as well as non-pregnant midlife women (ages 40 to 60), urinary phthalate biomarker concentrations were associated with altered serum and urinary sex-steroid hormone concentrations (86, 249). In addition to hormonal disruption, DEHP and its metabolites may disrupt other critical cellular processes, including cell viability and growth pathways, which may be responsible for the development of adverse health conditions in women, including uterine fibroids (611-613). For example, an in vitro study reported that cells isolated from human uterine fibroids and treated with DEHP had higher viability, lower 372 apoptosis, and increased expression of hypoxia inducible factor-1α and cyclooxygenase- 2 (613). Additionally, a cross-sectional study of pre-menopausal women found that urinary concentrations of one DEHP metabolite, mono-(2-ethyl-5-hydroxyhexyl) phthalate (MEHHP), were associated with altered micro-RNA levels (miR-10a-5p and miR-577), which are important for cell viability, survival, and apoptosis in fibroid tumors, suggesting that phthalates may play a role in fibroid pathogenesis by interacting with regulators of epigenetic machinery (611). Given this experimental evidence, additional studies in human populations are needed to determine whether these experimental findings recapitulate the real-life experience of women. Many observational studies have investigated associations between phthalate metabolite biomarker concentrations and uterine fibroids (83, 187, 608, 614). A 2017 meta-analysis found no significant pooled associations between total urinary phthalate metabolite concentrations and uterine fibroids, whereas DEHP metabolites were associated with higher odds of fibroids (608). Additional recent studies have confirmed positive associations between urinary DEHP metabolites, especially mono-2-ethylhexyl phthalate (MEHP) and fibroids (83, 614). A major limitation of these studies is that urine samples for phthalate biomarker assessment were collected after fibroid diagnosis, which makes it difficult to establish temporality and causality. A recent prospective study of 23–35-year- old premenopausal black women in Detroit that evaluated the associations between phthalates and ultrasound-detected fibroids (n=301) women who did and 453 who did not develop fibroids) found a weak-to-moderate association between MEHP and higher risk of fibroids (187), supporting results from previous studies. 373 Similar to most previous studies, our current study did not prospectively evaluate associations between phthalate biomarker concentrations and fibroids. However, our goal was to contribute additional findings on the overall associations between phthalate biomarkers and fibroids in a large, diverse cohort of pre- and peri-menopausal midlife women, and also propose potential effect modification by adulthood weight gain, which (to our knowledge) has not been previously considered. Our first objective was to evaluate the overall associations between urinary phthalate biomarker concentrations and prior fibroid diagnosis. As weight change, specifically weight gain (615), is a risk factor for fibroid development, our second objective was to evaluate if associations between phthalate biomarkers and prior fibroid diagnosis differed in women who became overweight/obese from age 18 to midlife compared to women whose body mass index (BMI) remained stable (remained under-/normal weight or overweight/obese). To improve the window of our exposure measure relative to the outcome, our sensitivity analyses also considered whether associations between phthalate biomarkers and prior fibroid diagnosis differed based on the timing of diagnosis relative to phthalate assessment. F.4. MATERIALS AND METHODS The current study was a secondary analysis of baseline data collected as part of the Midlife Women’s Health Study (MWHS). The MWHS is a prospective cohort that recruited midlife women living in and around Baltimore, Maryland, between 2006 and 2015, with the primary goal of assessing the risk factors of hot flashes. The study protocol has been described elsewhere (449). In brief, women were included in the study if they were between 45 and 54 years old and were pre- or peri-menopausal. Women were excluded 374 from the study if they had a history of hysterectomy or oophorectomy, were currently pregnant, were taking hormone therapy or herbal/other agents for menopause treatment, were taking oral contraceptives, were undergoing cancer treatment, or were postmenopausal. Menopause status was defined using the Stages of Reproductive Aging Workshop + 10 (STRAW+10) criteria as follows (558): premenopausal women were those who experienced their last menstrual period within the past three months and reported ≥ 11 periods within the past year; perimenopausal women were those who experienced their last menstrual period within the past year, but not within the past three months or experienced their last menstrual periods within the past three months and experienced ≤ 10 periods within the past year; and postmenopausal women were those who had not experienced a menstrual period within the past year. Overall, 754 pre- and peri- menopausal women with complete information about baseline midlife urinary phthalate metabolite concentrations and self-reported past uterine fibroid diagnosis were available for the study. At baseline, women reported their race/ethnicity, annual household income, alcohol intake, weight at age 18, oral contraceptive use, age at menarche, fertility consultation, and parity via a self-administered questionnaire. Women reported their race/ethnicity by selecting one of the following options: Caucasian/White, African American/Black, Hispanic, Asian, or other. To determine women’s most recent alcohol consumption status, women answered “yes” or “no” to the question, “In the last 12 months have you had at least 12 drinks of any kind of alcoholic beverage?”. Women reported “yes” or “no” if they ever used oral contraceptive pills. If they marked yes, they also reported the duration of 375 use. To evaluate if women had reproductive problems, they answered “yes” or “no” to the question “Did you ever seek medical consultation because of difficulty in getting pregnant (infertility)?”. At the first baseline clinic visit, trained staff measured women’s height (inches) and weight (pounds), which we used to calculate midlife BMI (kg/m2). We additionally used measured midlife height and self-reported weight at age 18 to calculate BMI at age 18 (kg/m2). Unfortunately, this study was not designed to assess fibroid incidence. Instead, the baseline questionnaire collected information about prior fibroid diagnosis. Women answered “yes” or “no” to the question “Have you ever been told by a doctor that you have uterine fibroids? If women reported “yes”, they indicated their age at diagnosis. We evaluated prior fibroid diagnosis as a binary variable comparing women who had a diagnosis to those without a diagnosis. We evaluated the timing of fibroid diagnosis as a three-level variable using the following categories: women who were diagnosed with fibroids within five years of the baseline visit, those diagnosed with fibroids five years or more before the baseline visit, and those never diagnosed with fibroids. We were unable to ascertain women’s phthalate exposure at the time of their fibroid diagnosis. Instead, at the baseline clinic visit, women provided a urine sample and then provided up to three more urine samples over consecutive weeks. Staff physically pooled urine samples for each participant for analysis of phthalate metabolite biomarkers, which were used to approximate midlife exposure to these chemicals. Given the short half-lives of phthalates in the body and high within-person variability, pooling has been shown to 376 be an effective approach for reducing measurement error in phthalate biomarkers (162). MWHS staff sent one pooled urine sample per participant to the University of Illinois Urbana-Champaign Roy K. Carver Biotechnology Metabolomics Center for analysis. The metabolomics laboratory used isotope dilution high-performance liquid chromatography negative-ion electrospray ionization-tandem mass spectrometry (HPLC–MS/MS) and methods adapted from the Centers for Disease Control and Prevention (104) to analyze urine samples for concentrations (ng/mL) of the following nine phthalate metabolites: MEHP, MEHHP, mono-(2-ethyl-5-carboxypentyl) phthalate (MECPP), mono-(2-ethyl- 5- oxohexyl) phthalate (MEOHP), mono-(3-carboxypropyl) phthalate (MCPP), monobenzyl phthalate (MBzP), monoethyl phthalate (MEP), monobutyl phthalate (MBP), and monoisobutyl phthalate (MiBP). We imputed phthalate metabolite concentrations below the limit of detection (LOD) using LOD/√2. All phthalate metabolite concentrations were specific gravity-adjusted to account for urine dilution using the following equation: Pc = P[(1.018 − 1)/(SG − 1)], where Pc is the specific gravity-adjusted metabolite concentration, P is the measured metabolite concentration (ng/mL), 1.018 is the median specific gravity of the MWHS population included in this analysis, and SG is the specific gravity of each woman’s pooled urine sample (109). We molar-converted and summed (mmol/mL) four DEHP metabolites (MEHP, MEHHP, MEOHP, and MECPP) to approximate exposure to DEHP (referred to as ΣDEHP). The remaining metabolites (MCPP, MBzP, MEP, MBP, and MiBP) were evaluated using non-molar converted concentrations (ng/mL). We created additional molar sums based on exposure sources, where metabolites MEHP, MEHHP, MEOHP, 377 MECPP, MCPP, and MBzP were molar summed to approximate exposure to plasticizer phthalates (ΣPlastics) and metabolites MEP, MBP, and MiBP were molar summed to approximate exposure to personal care product phthalates (ΣPCP). Some previous experimental studies in male pups suggest that metabolites MEHP, MEHHP, MEOHP, MECPP, MBzP, MBP, and MiBP have anti-androgenic activity in the body (149, 495, 616). Therefore, we molar summed these urinary biomarkers to approximate exposure to phthalates with anti-androgenic activity (ΣAA). Interestingly, in our previous study, we observed positive associations between ΣAA and estradiol (249), suggesting that classifying phthalate biomarkers based on their effects on male fetal rat testes may not be appropriate for evaluating reproductive endpoints in women. However, we evaluated associations between ΣAA and fibroids to corroborate findings from a previous study (614). Lastly, we molar summed all nine phthalate metabolites to approximate total midlife phthalate exposure (ΣPhthalates). We used the chi-squared test to evaluate differences in sociodemographic, lifestyle, and health characteristics between women with and without prior fibroid diagnosis. We then used Poisson regression models with robust variance estimator to evaluate associations between phthalate biomarker concentrations (as individual metabolites or molar sums) and prior fibroid diagnosis (617). Due to skewed distributions, phthalate biomarker concentrations were natural log-transformed. To address our first and second objectives, we specified Poisson regression models to evaluate associations between phthalate biomarker concentrations and the risk of having a prior fibroid diagnosis compared to not having a prior fibroid diagnosis. To evaluate differences in associations between 378 phthalate biomarker concentrations and prior fibroid diagnosis by changes in BMI from age 18 to 45-54 (second objective), we first classified both midlife BMI and BMI at age 18 using the following clinical categories (618): under-/normal weight (< 25 kg/m2) and overweight/obese (≥ 25 kg/m2). Then, we categorized changes in BMI as follows: women who remained overweight/obese through age 45-54 (overweight/obese at ages 18 and 45-54), women who became overweight/obese by age 45-54 (under-/normal weight at age 18 but overweight/obese at age 45-54), women who remained under-/normal weight through age 45-54 (under-/normal at ages 18 and 45-54), and those who became under- /normal weight by age 45-54 (overweight/obese at age 18 but under-/normal weight at age 45-54) (248). In these models, we included a multiplicative interaction between phthalates and change in BMI to evaluate differences in associations in women who remained overweight/obese through age 45-54, who became overweight/obese by age 45-54, and who remained under-/normal weight. We excluded women who became under-/normal weight because this category only included five women. We reported results regardless of the significance of the interaction P-value. We acknowledge that our study was not designed to prospectively evaluate the associations between urinary phthalate biomarker concentrations and fibroids diagnosis, and that midlife urinary phthalate biomarker concentrations likely do not represent concentrations at the time of prior fibroid diagnosis. However, in addition to our main analyses, we also conducted a sensitivity analysis to potentially provide a more relevant approximation of phthalate exposure in relation to prior fibroid diagnosis. Specifically, we wanted to assess whether our primary associations differed based on when women were 379 diagnosed with fibroids (timing) relative to when they provided their midlife urine samples for phthalate metabolite quantification. We used multinomial logistic regression models to evaluated the associations between urinary phthalate biomarker concentrations and the probability of being diagnosed with fibroids within five years of or more than five years before midlife urine collection compared to never being diagnosed with fibroids. For our first objective, we assessed both unadjusted and adjusted models. In adjusted models, we a priori selected covariates associated with both our exposure and outcome. We evaluated correlations between all selected covariates to test for potential multicollinearity issues, but none were strongly correlated with each other (r < 0.4; data not shown). Therefore, final adjusted models included the following covariates: race/ethnicity, annual household income, alcohol intake, fertility consultation, midlife BMI, oral contraceptive use, age at menarche, and parity. These covariates are proxies of important latent constructs that we were unable to directly assess at the time of fibroid diagnosis, such as socioeconomic status (race/ethnicity, income), racism (race/ethnicity), lifestyle (midlife BMI, alcohol use), health (midlife BMI, alcohol use, age at menarche, fertility consultation), and reproductive history (oral contraceptive use, parity, age at menarche, fertility consultation). These covariates were also accounted for in sensitivity analyses. For our second objective, we included the previously listed covariates, except for midlife BMI due to multicollinearity issues with changes in BMI. The operationalization of covariates and reference groups are presented in Table 43. 380 We conducted all analyses in SAS 9.4 (SAS Institute Inc, Cary, NC, USA). We used PROC GENMOD for Poisson regression analyses with a robust variance estimator (main analyses) and specified an unstructured correlation matrix for the model’s residuals. We back-transformed the resulting risk ratios (RR) and 95% confidence intervals (CI) using the equation [exp(ln(RR)*ln(2.00)] to interpret the results as risk of prior fibroid diagnosis with every two-fold increase in phthalate biomarker concentration. We used PROC LOGISTIC for multinomial logistic regression models (sensitivity analysis), and back- transformed the resulting odds ratios (OR) and 95% CIs using the equation [exp(ln(OR)*ln(2.00)] to interpret the results as odds of being diagnosed with fibroids within five years or more than five years before midlife urine collection with every two-fold increase in phthalate biomarker concentration. We considered associations as being meaningful at P ≤ 0.10 and analyses were not adjusted for multiple comparisons (166). F.5. RESULTS F.5.1. Demographic and lifestyle characteristics of the MWHS population The characteristics of the overall MWHS population have been described elsewhere (249, 250). The prevalence of prior fibroid diagnosis in MWHS was approximately 27% (Table 43), and median age at diagnosis was 40 years (range: 16 – 52; data not shown). Women with and without prior fibroid diagnosis differed significantly with regard to race/ethnicity, annual household income, alcohol intake, midlife BMI, age at menarche, oral contraceptive use, fertility consultation, and parity (P < 0.05; Table 43). Specifically, compared to women with no prior fibroid diagnosis, women with prior diagnosis were more likely to be black/other, have lower annual household incomes, have obesity during 381 midlife, start menarche earlier, have no or one live birth, use oral contraceptives for >10 years, consume ≤ 12 alcoholic drinks in the year before the first study visit, and not to have received fertility treatment when trying to become pregnant. Table 43. Demographic and lifestyle characteristics of women with and without uterine fibroids. Fibroid diagnosis Participant characteristic Yes (n=207) No (n=547) P-value Age at baseline 0.49 45 – 49 years 131 (63.3) 361 (66.0) 50 – 54 years 76 (36.7) 186 (34.0) Race/ethnicity < 0.0001 Non-Hispanic white 88 (42.7) 407 (74.5) Black/other1 118 (57.3) 139 (25.5) Employment 0.94 Unemployed 41 (19.9) 110 (20.1) Employed 165 (80.1) 436 (79.9) Educational attainment 0.25 Some college or less 79 (38.5) 186 (34.1) College graduate or higher 126 (61.5) 360 (65.9) Annual household income 0.01 < $20,000 14 (7.0) 35 (6.6) $20,000 – 39,999 41 (20.4) 78 (14.7) $40,000 – 99,999 78 (38.8) 169 (31.8) ≥ $100,000 68 (33.8) 249 (46.9) Marital status 0.21 Single 42 (20.5) 92 (16.8) Married/living with partner 124 (60.5) 368 (67.4) Widowed/divorced/separated 39 (19.0) 86 (15.8) Alcohol intake 0.001 No 91 (44.2) 169 (31.0) Yes 115 (55.8) 376 (69.0) Ever smoker 0.31 Yes 86 (41.7) 251 (45.9) No 120 (58.3) 296 (54.1) Menopause status 0.61 Premenopausal 137 (66.2) 351 (64.2) Perimenopausal 70 (33.8) 196 (35.8) Midlife BMI 0.02 < 25 kg/m2 70 (33.8) 230 (42.0) 25 – 29.9 kg/m2 52 (25.1) 149 (27.2) ≥ 30.0 kg/m2 85 (41.1) 168 (30.7) Age at menarche 0.001 < 12 years 119 (58.0) 231 (42.5) 13 – 14 years 64 (31.2) 238 (43.8) ≥ 15 years 22 (10.7) 74 (13.6) 382 Table 43 (cont’d). Fibroid diagnosis Participant characteristic Yes (n=207) No (n=547) P-value Oral contraceptive use 0.04 Never 24 (11.7) 86 (15.8) < 1 year 35 (17.1) 70 (12.8) 1 – 4 years 51 (24.9) 146 (26.8) 5 – 10 years 40 (19.5) 139 (25.5) > 10 years 55 (26.8) 104 (19.1) Fertility consultation 0.03 Yes 28 (13.7) 111 (20.4) No 177 (86.3) 433 (79.6) Parity 0.03 Never pregnant 20 (9.7) 66 (12.1) No live births 25 (12.1) 50 (9.2) 1 live birth 49 (23.8) 86 (15.8) ≥ 2 live births 112 (54.4) 344 (63.0) Change in BMI from age 18 to 45-54 0.29 Remained under-/normal weight 68 (33.7) 224 (41.2) Became overweight/obese 107 (53.0) 253 (46.5) Became under-/normal weight 1.0 (0.5) 4 (0.7) Remained overweight/obese 26 (12.9) 63 (11.6) 1Women of other race/ethnicity represent less than 4% of the analytic sample. P-value from chi- squared test. Missing information (n) for women with prior fibroid diagnosis: race/ethnicity, employment, alcohol intake, ever smoker, parity (n=1); education, marital status, age at menarche, oral contraceptive use, fertility consultation (n=2); change in BMI from age 18 to 45-54 (n=5); income (n=6). Missing information (n) for women without prior fibroid diagnosis: race/ethnicity, employment, education, marital status, parity (n=1); alcohol intake, oral contraceptive use (n=2); fertility consultation, change in BMI from age 18 to 45-54 (n=3); age at menarche (n=4); income (n=16). BMI, body mass index. F.5.2. Urinary phthalate metabolite biomarker concentrations More than 99% of women had urinary concentrations of all measured phthalate metabolites above the LOD (data not shown). MWHS women had somewhat higher median phthalate metabolite concentrations compared to midlife women from the National Health and Nutrition Examination Survey (NHANES) for the years 2005-2016, but with overlapping 25th and 75th percentiles (249, 250) (Figure 30). 383 Figure 30. Midlife urinary phthalate metabolite concentrations. Box plots display urinary phthalate metabolite concentrations (ng/mL) of women in MWHS (2006-2015, n=754) and women ages 45-54 from 6 NHANES survey cycles (2005-2016, n=902). Concentrations were not adjusted for urine dilution. Box plots include the median (center line in box), the 25th percentile (lower line of box), and the 75th percentile (upper line in box). Numeric values for phthalate metabolite concentrations have been published elsewhere (249, 250). MEHP (mono-2-ethylhexyl phthalate); MEHHP (mono-(2-ethyl-5- hydroxyhexyl) phthalate); mono-(2-ethyl- 5-oxohexyl) phthalate (MEOHP); mono-(2- ethyl-5-carboxypentyl) phthalate (MECPP); mono-(3-carboxypropyl) phthalate (MCPP); monobenzyl phthalate (MBzP); monoethyl phthalate (MEP); monobutyl phthalate (MBP); and monoisobutyl phthalate (MiBP); MWHS, Midlife Women’s Health Study; NHANES, National Health and Nutrition Examination Survey. F.5.3. Associations of phthalate biomarker concentrations with prior fibroid diagnosis Overall, higher concentrations of some phthalate biomarkers were associated with higher risk of prior fibroid diagnosis (Table 44). In unadjusted models, women had 10% - 19% higher risk of prior fibroid diagnosis for every two-fold increase in ΣDEHP (RR: 1.10; 95% CI: 1.00, 1.22), MEP (RR: 1.11; 95% CI: 1.02, 1.20), MiBP (RR: 1.18; 95% CI: 1.06, 1.31), ΣPCP (RR: 1.16; 95% CI: 1.08, 1.25), ΣPhthalates (RR: 1.19; 95% CI: 1.10, 1.29), and ΣAA (RR: 1.15; 95% CI: 1.02, 1.29). Interestingly, in unadjusted models, we also 384 observed a marginal inverse association of MBzP with prior fibroid diagnosis (RR: 0.92; 95% CI: 0.84, 1.02). After accounting for important confounders, only associations of ΣDEHP, ΣPhthalates, and ΣAA with fibroids remained. Women had a 9-16% higher risk of prior fibroid diagnosis, respectively, for every two-fold increase in ΣDEHP (RR: 1.13; 95% CI: 1.02, 1.26), ΣPhthalates (RR: 1.09; 95% CI: 1.00, 1.19), or ΣAA (RR: 1.16; 95% CI: 1.03, 1.31). In adjusted models, we also observed a marginal association with ΣPlastics, where women had a 12% (RR: 1.12; 95% CI: 1.00, 1.25) higher risk of prior fibroid diagnosis for every two-fold increase in ΣPlastics. Table 44. Associations of phthalate biomarker concentrations with diagnosis of uterine fibroids. Unadjusted (n=754) Adjusted (n=712) Phthalate biomarker RR (95% CI) RR (95% CI) ∑DEHP1 1.10 (1.00, 1.22)# 1.13 (1.02, 1.26)* MCPP 0.99 (0.92, 1.07) 0.99 (0.91, 1.07) MBzP 0.92 (0.84, 1.02)# 0.97 (0.88, 1.07) ∑Plastics2 1.08 (0.97, 1.20) 1.12 (1.00, 1.25)# MEP 1.11 (1.02, 1.20)* 1.01 (0.94, 1.09) MBP 1.06 (0.94, 1.20) 1.05 (0.94, 1.19) MiBP 1.18 (1.06, 1.31)* 1.09 (0.98, 1.22) ∑PCP3 1.16 (1.08, 1.25)* 1.05 (0.97, 1.14) ∑Phthalates4 1.19 (1.10, 1.29)* 1.09 (1.00, 1.19)* ∑AA5 1.15 (1.02, 1.29)* 1.16 (1.03, 1.31)* Poisson regression models with robust variance estimator evaluating the risk of being diagnosed with fibroids (unadjusted model n=207, adjusted model n=193) compared to never being diagnosed with fibroids (unadjusted model n=547, adjusted model n=519) for every 2-fold increase in phthalate biomarker concentration. Adjusted models account for race/ethnicity, income, age at menarche, oral contraceptive use, parity, fertility consultation, and midlife BMI. 1∑DEHP = MEHP/278 + MEHHP/294 + MEOHP/292 + MECPP/308; 2∑Plastics = MEHP/278 + MEHHP/294 + MEOHP/292 + MECPP/308 + MCPP/252 + MBzP/256; 3∑PCP = MEP/194 + MBP/222 + MiBP/222; 4∑Phthalates = MEHP/278 + MEHHP/294 + MEOHP/292 + MECPP/308 + MCPP/252 + MBzP/256 + MEP/194 + MBP/222 + MiBP/222; 5∑AA = MEHP/222 + MEHHP/294 + MEOHP/292 + MECPP/308 + MBzP/256 + MBP/222 + MiBP/222. BMI, body mass index. CI, confidence interval; RR, risk ratio. #P ≤ 0.10 and *P < 0.05. 385 F.5.4. Differences in associations by change in BMI from 18 years of age to midlife Associations of phthalate biomarker concentrations with prior fibroid diagnosis were strongest in women who became overweight or obese (Figure 31). Specifically, women who became overweight or obese had a 17 - 21% higher risk of having a prior fibroid diagnosis for every two-fold increase in ΣDEHP (RR: 1.21; 95% CI: 1.06, 1.38), ΣPlastics (RR: 1.17; 95% CI: 1.01, 1.35), and ΣAA (RR: 1.20; 95% CI: 1.04, 1.39). Additionally, we observed a marginal association between ΣPhthalates and fibroids, where women who became overweight or obese had a 10% (RR: 1.10; 95% CI: 1.00, 1.22) higher risk of having a prior fibroid diagnosis for every two-fold increase in ΣPhthalates. Lastly, we observed a marginal association between ΣAA and fibroids, where women who remained under-/normal weight had a 22% (RR: 1.22; 95% CI: 0.96, 1.55) higher risk of having a prior fibroid diagnosis for every two-fold increase in ΣAA. 386 Figure 31. Differences in associations of phthalate biomarker concentrations with uterine fibroids by change in BMI from age 18 to 45-54. Forest plots display risk ratios (filled circle) and 95% confidence intervals (horizonal lines) for the risk of prior fibroid diagnosis for every 2-fold increase in phthalate biomarker concentrations among women who remained overweight/obese (n=223 no prior fibroids diagnosis, n=66 yes prior fibroids diagnosis), became overweight/obese (n=135 no prior fibroids diagnosis, n=47 yes prior fibroids diagnosis), and remained under-/normal weight (n=161 no prior fibroids diagnosis, n=80 yes prior fibroids diagnosis) from age 18 to 45-54. Models account for race/ethnicity, income, age at menarche, oral contraceptive use, parity, fertility consultation, change in BMI from age 18 to 45-54, as well as a multiplicative interaction between phthalate biomarker and change in BMI. Confidence intervals that do not cross the null (dashed vertical line) are significant at #P ≤ 0.10 or *P < 0.05. ∑DEHP = MEHP/278 + MEHHP/294 + MEOHP/292 + MECPP/308; ∑Plastics = MEHP/278 + MEHHP/294 + MEOHP/292 + MECPP/308 + MCPP/252 + MBzP/256; ∑PCP = MEP/194 + MBP/222 + MiBP/222; ∑Phthalates = MEHP/278 + MEHHP/294 + MEOHP/292 + MECPP/308 + MCPP/252 + MBzP/256 + MEP/194 + MBP/222 + MiBP/222; ∑AA = MEHP/222 + MEHHP/294 + MEOHP/292 + MECPP/308 + MBzP/256 + MBP/222 + MiBP/222. F.5.5. Associations of phthalate biomarker concentrations in women with more recent diagnosis Among women who had a prior fibroids diagnosis, median time of diagnosis relative to midlife urine collection for phthalate biomarker assessment was 8 years (range: 0 – 33 387 years). In sensitivity analyses, the associations between some phthalate biomarkers and fibroid diagnosis differed according to the recency of diagnosis relative to urine collection for phthalate biomarker assessment (Table 45). Overall associations between ΣDEHP, ΣPlastics, ΣPhthalates, ΣPCP, and ΣAA and prior fibroid diagnosis were more robust in women diagnosed within five years of midlife urine collection. Specifically, women diagnosed within five years of midlife urine collection generally had 19% - 35% higher odds of prior fibroid diagnosis for every two-fold increase in ΣDEHP (OR: 1.29; 95% CI: 1.03, 1.61), ΣPlastics (OR: 1.23; 95% CI: 0.96, 1.56), ΣPCP (OR: 1.30; 95% CI: 0.99, 1.43), ΣPhthalates (OR: 1.30; 95% CI: 1.05, 1.61), and ΣAA (OR: 1.35; 95% CI: 1.04, 1.76). Table 45. Associations of urinary phthalate biomarker concentrations with timing of uterine fibroid diagnosis. Fibroid diagnosis ≥ 5 years Fibroid diagnosis < 5 years before midlife urine collection before midlife urine collection (n=111) (n=82) Phthalate biomarker OR (95% CI) OR (95% CI) ∑DEHP1 1.18 (0.95, 1.45) 1.29 (1.03, 1.61)* MCPP 0.96 (0.83, 1.11) 1.01 (0.86, 1.18) MBzP 1.06 (0.87, 1.28) 0.84 (0.68, 1.05) ∑Plastics2 1.17 (0.93, 1.46) 1.23 (0.96, 1.56)# MEP 0.98 (0.86, 1.11) 1.08 (0.93, 1.26) MBP 1.02 (0.80, 1.31) 1.16 (0.91, 1.49) MiBP 1.14 (0.91, 1.44) 1.16 (0.91, 1.49) ∑PCP3 1.00 (0.84, 1.20) 1.19 (0.99, 1.43)# ∑Phthalates4 1.06 (0.86, 1.33) 1.30 (1.05, 1.61)* ∑AA5 1.20 (0.93, 1.55) 1.35 (1.04, 1.76)* Multinomial logistic regression models evaluated the odds of being diagnosed with fibroids ≥ 5 years (n=111) or < 5 years (n=82) before baseline compared to never being diagnosed with fibroids (n=519) for every 2-fold increase in phthalate biomarker concentration. Models account for race/ethnicity, income, age at menarche, oral contraceptive use, parity, fertility consultation, and midlife body mass index. 1∑DEHP = MEHP/278 + MEHHP/294 + MEOHP/292 + MECPP/308; 2∑Plastics = MEHP/278 + MEHHP/294 + MEOHP/292 + MECPP/308 + MCPP/252 + MBzP/256; 3∑PCP = MEP/194 + MBP/222 + MiBP/222; 4∑Phthalates = MEHP/278 + MEHHP/294 + MEOHP/292 + MECPP/308 + MCPP/252 + MBzP/256 + MEP/194 + MBP/222 + MiBP/222; 5∑AA = MEHP/222 + MEHHP/294 + MEOHP/292 + MECPP/308 + MBzP/256 + MBP/222 + MiBP/222. CI, confidence interval; OR, odds ratio. #P ≤ 0.10 and *P < 0.05. 388 F.6. DISCUSSION Overall, we observed that ΣDEHP was associated with a higher risk of prior fibroid diagnosis, which was the main contributor to the associations between ΣPlastics, ΣPhthalates, and ΣAA and prior fibroid diagnosis. These associations were strongest in women who became overweight/obese from ages 18 to 45-54. The associations between phthalate biomarker concentrations and prior fibroid diagnosis were also stronger in women diagnosed with fibroids within five years of midlife urine collection for phthalate biomarker assessment. Our overall results related to ΣDEHP with prior fibroid diagnosis corroborate those from previous studies. However, additional large-scale prospective studies in diverse populations are needed. Our findings that ΣDEHP and related phthalate molar sums are associated with higher odds of prior fibroid diagnosis are consistent with previous experimental and observational studies. Specifically, experimental studies observed that human fibroid cells treated with DEHP metabolites had disrupted cell viability, apoptotic, and growth pathways (611-613). A 2017 meta-analysis pooled five observational studies conducted between 1999 and 2015 in populations from the U.S., Korea, China, and Taiwan, and observed that DEHP metabolites were associated with higher odds of fibroids (608). However, this meta-analysis also reported that MBP and MiBP were also marginally significantly associated with higher odds of fibroids. These differences in findings may be related to the study population, since women in the meta-analysis were predominately non-Hispanic white or Asian, while most women in our population were non-Hispanic white or black. Additionally, our study population included women between the ages of 389 45 and 54, while the meta-analysis included a wider age range of women. Our findings are also consistent with a recent U.S. cross-sectional study of predominately black 26 - 54-year-old women (recruited 2014 - 2017) undergoing hysterectomy or myomectomy that reported positive associations between ΣDEHP and ΣAA and fibroid volume (614). Similarly, a case-control study of 20 - 40-year-old Korean women (recruited 2015 - 2016) reported that the odds of fibroids was higher in women in quartiles 2, 3, and/or 4 (compared to quartile 1) of urinary concentrations of ƩDEHP and its metabolites (83). These two more recent studies collected spot urine samples for phthalate biomarker assessment and ascertained fibroid status using imaging technology (i.e., magnetic resonance imaging (MRI), ultrasound) and/or pathology reports (83, 614). Despite the differences in study population, as well as phthalate biomarker and fibroid assessment, in these studies compared to ours, associations between DEHP and its metabolites and fibroids appear to be consistent. The causal interpretability of results from most studies evaluating the associations of phthalate metabolite biomarkers with fibroids is limited because urine collection for phthalate biomarker assessment often occurred after women had already been diagnosed with fibroids. To our knowledge, only one study of Detroit-area 23 – 35-year- old black women (recruited 2010 - 2012) prospectively evaluated associations between phthalate biomarkers and incidence of fibroids (187). Although we were also unable to prospectively evaluate these associations prospectively, we conducted sensitivity analyses to address this limitation. In sensitivity analyses assessing the timing of fibroid diagnosis relative to midlife urine collection for phthalate biomarker assessment, we found 390 that associations of ΣDEHP and related phthalate molar sums with higher likelihood of prior fibroid diagnosis were stronger in women diagnosed within five years of urine collection. This five-year cutoff makes it more likely that women’s urinary phthalate biomarker concentrations at the time of fibroid diagnosis were similar to those during the study. However, even with these sensitivity analyses, we cannot rule out the potential of reverse causation. Women diagnosed with fibroids may have developed certain lifestyles or behaviors over the years leading up to their enrollment into the MWHS that may have influenced their midlife exposure to phthalates. Additionally, given the high temporal variability in urinary phthalate biomarker concentrations (619), we are likely not capturing the correct exposure window, which may result in the underestimation of associations between phthalate biomarkers and risk of fibroids. Therefore, additional prospective studies are needed to elucidate the directionality of associations between phthalate biomarkers and fibroid development, and to consider the sensitive window of exposure framework to establish the temporal relationship between phthalate exposure and fibroid development. Nevertheless, the findings from this study should be interpreted with caution. To our knowledge, our study is the first to show that associations between ΣDEHP and related phthalate molar sums and prior fibroid diagnosis were driven by women who became overweight or obese from age 18 to midlife. Obesity and especially weight gain are important risk factors for the development of fibroids. A recent meta-analysis that pooled 22 studies observed that higher body weight and adiposity (measured by BMI, waist/hip ratio, and waist circumference), as well as weight gain since age 18 were 391 associated with higher odds of fibroids (620). Given that both phthalates and gaining weight since age 18 are determinants of fibroid development, our findings suggest that women who undergo major changes in weight may be more susceptible to the impact of phthalates than those whose weights remain stable from age 18 until midlife. These results could be due to the metabolic disruptions that occur with weight change. For example, changes in weight/fat distribution and reproductive hormones (i.e., estrogens) are linked, with obesity influencing hormone concentrations (440, 475), as well as reproductive hormones influencing changes in weight and fat deposition (434). Given this relationship between body composition/adiposity and hormones, the interaction between endocrine disrupting chemicals (i.e., phthalates) and adipose tissue may contribute to the increased risk of phthalate-induced fibroids in women who gained weight. While most individuals experience gradual weight gain across the lifespan (621), because we did not have information about women’s weight at fibroid diagnosis, it is possible that some weight gain may have occurred after diagnosis. Additionally, women who gained weight may have engaged in certain behaviors or have lifestyles resulting in the use of products associated with increased exposure to DEHP, which may explain why associations between ΣDEHP and related phthalate molar sums and prior fibroid diagnosis were stronger in these women. For example, unhealthy dietary behaviors, such as consumption of processed and fast foods, are major determinants of phthalate exposure (240). A study using data from NHANES found that individuals who consumed more fast foods had higher urinary ƩDEHP concentrations than non-consumers (316). Future studies that consider diet are needed to corroborate our findings, and to address 392 biologically plausible pathways connecting phthalate exposure and weight change with the development of fibroids. This current study has important strengths, but also some limitations in addition to the potential for reverse causation. First, we may be underpowered to detect certain associations of phthalate biomarkers with prior fibroid diagnosis, especially in sensitivity analyses and in analyses evaluating differences by changes in BMI from age 18 to midlife. However, we evaluated these cross-sectional associations in a large cohort of pre- and peri-menopausal midlife women with a relatively high prevalence of fibroids. Second, prior fibroid diagnosis was based on self-reports and women with a history of hysterectomy or oophorectomy were excluded, which could lead to the misclassification of fibroid status and underestimation of the number of women with fibroids (622). However, our findings that race/ethnicity, BMI, and age at menarche are important predictors of self-reported fibroid diagnosis are consistent with the prior literature (623). Additionally, our results related to the associations between ƩDEHP and prior fibroid diagnosis are also consistent with previous studies that determined fibroid status using imaging technology (i.e., MRI, ultrasound) and/or pathology reports (83, 614). Third, we used self-reported weight at age 18 to calculate changes in BMI from age 18 to midlife, which may also be subject to recall bias. However, at the population level, self-reported past body weight is reliable for predicting measured past body weight (478). Fourth, we are unable to causally interpret our results given that urine for midlife phthalate biomarker assessment was collected after women’s fibroid diagnosis. However, we conducted sensitivity analyses and observed that associations between phthalate biomarkers and fibroids are stronger when the time 393 between midlife urine collection and prior fibroid diagnosis is reduced. Additionally, phthalate metabolite concentrations were quantified from pools of up to four urine samples, which provides a more stable measure of our exposure and may better represent midlife urinary concentrations (162, 412). Lastly, as previously discussed, there may be unmeasured confounding factors (i.e., by diet), which were unaccounted for in our statistical models evaluating associations between phthalate biomarkers and prior fibroid diagnosis. However, we selected covariates a priori using the previous literature, and included covariates that are proxies for important latent constructs that we were unable to directly assess at the time of fibroid diagnosis. Overall, this secondary data analysis leverages data collected as part of the MWHS to contribute additional information pertaining to associations between phthalate biomarkers and prior fibroid diagnosis in midlife women. F.7. CONCLUSION In this population of mostly non-Hispanic white or black pre- and peri-menopausal midlife women, we observed that ΣDEHP and related phthalate molar sums were associated with higher risk of prior fibroid diagnosis, and these associations were stronger in women who became overweight or obese from age 18 to midlife. Our findings corroborate results from previous experimental and observational studies and suggest that interventions targeting the lifestyle behaviors associated with phthalate exposure are needed. However, as with most previous studies, we were unable to prospectively evaluate these associations, which makes it challenging to causally interpret our results, and our findings should be interpreted with caution. Therefore, future longitudinal, prospective cohort 394 studies are needed to corroborate these findings and further elucidate the independent and interactive contribution of phthalates and weight gain to the development of uterine fibroids. 395