University of Southern California Dissertations and Theses

Prenatal environmental exposures and fetal growth in the MADRES cohort

(USC Thesis Other)

Prenatal environmental exposures and fetal growth in the MADRES cohort

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content PRENATAL ENVIRONMENTAL EXPOSURES AND FETAL GROWTH IN THE MADRES COHORT by Alicia K. Peterson A Dissertation Presented to the FACULTY OF THE USC GRADUATE SCHOOL UNIVERSITY OF SOUTHERN CALIFORNIA In Partial Fulfillment of the Requirements for the Degree DOCTOR OF PHILOSOPHY (EPIDEMIOLOGY) August 2022 Copyright 2022 Alicia K. Peterson ii ACKNOWLEDGMENTS A completed dissertation is in no way accomplished alone, and I am forever grateful to those who have played an instrumental role within my graduate studies. During this endeavor, there have been many who have provided encouragement and guidance, but the following are the ones I am especially grateful to. To my dissertation chair, Dr. Carrie Breton and my additional primary mentor, Dr. Theresa Bastain: My full appreciation for your impact on my personal and professional development cannot be properly put into words. Thank you for always believing in me, encouraging me to reach for the next step, and mentoring me through the last six years while being a research intern, then research staff, and finally a graduate student. I am incredibly grateful for your consistent advocacy and guidance in fostering a strong research foundation for my future endeavors. You both constantly inspire me, and this feat would not have been possible without your leadership, feedback, and support. “Thank You” will never be enough. To my dissertation committee members: Dr. Rima Habre, thank you for sharing your expertise in exposure science, motivating me to consider different perspectives, teaching me countless times about SAS code and JMP techniques, and always being my cheerleader. Dr. Sandrah Eckel, thank you for being the statistical queen that you are and for always being available to provide feedback and discuss methodological approaches; you introduced me to data analysis, and I will never forget the skills I learned from your 511a and 511c courses. I am so grateful for your guidance, encouragement, and teaching approach. Dr. Brendan Grubbs, thank you for being iii a soundboard during my presentations, helping me confirm biological plausibility, and always encouraging me throughout this process. To my very first mentor, Dr. Claudia Toledo-Corral: Thank you for introducing me to research, encouraging me to chase bigger dreams, and for always embracing me for who I am. I cherish your presence in my life. To my personal village: My mother, Beverley Peterson, who never fails to remind me how proud she is and for always celebrating the milestones; my partner, Dr. Kush Gaur, who has been by my side for the duration of this process and who always makes good days great and bad days better; my dearest friends, Mary Goral, Jackie Santana, Remy Landon, Ixel Hernandez-Castro, Marissa Kennerson, Molly Marshall, Jen Suh, Hannah Lee, Elizabeth Vang, Dr. Deborah Hoe, and soon to be Dr. Gabriela Toledo who all have cheered me on and encouraged me to do hard things. Lastly, a special acknowledgement to: Sally Peterson, Jane Hirschkowitz, and Dr. Madeline Eccleston. To my fellow PhD students: Ixel Hernandez-Castro, thank you for being my everything- buddy, offering countless proofreads, and always empathizing with the trials and tribulations. Christine Naya, thank you for being a student to look up to as an amazing example and for always being so open to share your work, tricks-of-the-trade, and advice throughout this process. Sahra Mohazzab-Hosseinian, thank you for these past years of encouragement, support, thoughtful feedback, and for always having tea or snacks on hand to share. An additional thank you to fellow students Karl O’Sharkey and Yan Xu for being a source of support, your consistent encouragement, and for helping me prepare for my defense. iv To those within the Department of Population and Public Health Sciences who have made a special impact on me through this process: Steve Howland, Vivien Le, Josie Tadeo, Milena Amadeus, Lisa Valencia, Carmen Chavez, Mary Trujillo, Renee Stanley, Sherri Fagan, Leticia Gracia, Celia Cedillo, Dr. Ugonna Ihenacho, and Dr. Zhongzheng (Jason) Niu. A notable acknowledgement to the following faculty within the department: Dr. Roberta McKean-Cowdin, Dr. Victoria Cortessis, Dr. Wendy Mack, and Dr. Kimberly Siegmund. I would also like to thank additional co-authors on my dissertation work who have contributed in multiple capacities: Dr. Zhongzheng Niu, Dr. Shohreh Farzan, Tingyu Yang, Dema Faham, Monica Amin, Dr. Deborah Lerner, Fred Lurmann, Nathan Pavlovic, Dr. Daphne Walker, Dr. Laila A. Al-Marayati, Dr. Edward Grant, Dr. Kurunthachalam Kannan and Morgan Robinson. Last, but certainly not least, none of this would have been possible without the participants of the MADRES study who have graciously contributed their time and information, the MADRES investigator team, the MADRES Staff and Data team, and our community clinic partners. v TABLE OF CONTENTS ACKNOWLEDGMENTS………………………………………………...………………………ii LIST OF TABLES………………………………………………………...………….…………viii LIST OF FIGURES………………………………………………………………………….……ix ABSTRACT……………………………………………………………………………………...xi OVERVIEW……………………………………………………………………………………....1 CHAPTER ONE…………………………………………………………………………………..6 Introduction to Per-and Polyfluoroalkyl Substances (PFAS)……………..……….…........6 Literature Review of PFAS on Fetal Growth and Birth Outcomes…………..……………...9 Biological Mechanisms for how Prenatal PFAS Exposure Influences Fetal Growth……9 Prenatal PFOA Exposure………………………………………………………….........11 Birthweight……………………………………………………………………………………..11 Gestational Age at Birth………………………………………………………………..........15 In Utero Fetal Growth……………………………………………………………...………..16 Prenatal PFOS Exposure……………………………………………………………......17 Birthweight……………………………………………………………………………………..17 Gestational Age at Birth………………………………………………………………...........20 In Utero Fetal Growth……………………………………………………………………….20 Prenatal PFHxS Exposure………………………………………………………….......22 Birthweight…………………………………………………………………………………....22 Gestational Age at Birth…………………………………………………………….............26 In Utero Fetal Growth………………………………………………………………………27 Prenatal PFNA Exposure………………………………………………………….........28 Birthweight……………………………………………………………………………………28 Gestational Age at Birth………………………………………………………………..........33 In Utero Fetal Growth……………………………………………………………………….34 Prenatal PFDA Exposure…………………………………………………………….....35 Birthweight……………………………………………………………………………………35 Gestational Age at Birth………………………………………………………………..........38 In Utero Fetal Growth……………………………………………………………………….39 Gaps Within the Literature……………………………………………………………….39 CHAPTER TWO………………………………………………………………………………...42 Introduction to Ambient Air Pollution……………………...…………………………....42 Ambient Air Pollution and Ultrasound Measures of Fetal Growth……………………...44 Biological Mechanisms – Prenatal Ambient Air Pollution and Fetal Growth……........45 Prenatal NO2 Exposure and Fetal Growth……..……………………………………..........46 Prenatal PM10 Exposure and Fetal Growth…………..………….…….………….......49 Prenatal PM2.5 Exposure and Fetal Growth…………..…………………………………….52 Prenatal O3 Exposure and Fetal Growth…………...……………..…………………..........55 vi Gaps Within the Literature………………………………………………………………56 CHAPTER THREE……………………………………………………………………………...59 Overview of the MADRES Study………………………….…………………………….59 Cohort Description…………………………..………………………………………..59 Relevant Methodology…………………………………………………………………...61 Prenatal PFAS Blood Serum Concentration…………………………………...……...61 Prenatal Ambient Air Pollution Exposure……………………………………………..62 Infant Birthweight and Gestational Age at Birth………………………………………63 In Utero Fetal Growth Parameters………………………………………………...…..63 Covariates.……………………………………………………………………………...64 MADRES Funding Sources ……………………………………………………………..64 CHAPTER FOUR Prenatal PFAS Exposure and Birth Outcomes…………………………………………………...65 Abstract………………………………………………………………………………..…65 Introduction………………………………………………………………………………66 Methods…………………………………………………………………………………..69 Study Sample…………………………………………………………………………...69 Exposure Assessment of Prenatal PFAS Concentrations………………………………....70 Birth Outcomes…………………………………………………………………………............72 Covariates………………………………………………………………………………............72 Statistical Analysis……………………………………………………………………………..74 Results……………………………………………………………………………………76 Participant Characteristics……………………………………………………………………76 PFAS Analytes………………………………………………………………………………….77 Single Pollutant Linear Regression Models………………………………………………...79 Multipollutant Linear Regression Models…………………………………………………..80 Sensitivity Analyses…………………………………………………………………………...81 Discussion………………………………………………………………………………..81 Conclusion…………………………………………………………………………….....85 Supplemental Material…………………………………………………………………...87 CHAPTER FIVE Prenatal PFAS Exposure and In Utero Fetal Growth……………………………………………92 Abstract………………………………………………………………………………..…92 Introduction……………………………………………………………………………....94 Methods………………………………………………………………………………..…96 Study Sample……………………………………………………………………………………96 Serum PFAS Analysis……………………………………………………………………….....97 Fetal Growth Assessment………………………………………………………………..........99 Perceived Stress Scale………………………………………………………………………..100 Covariates……………………………………………………………………………………...100 Statistical Analysis……………………………………………………………………………102 Results…………………………………………………………………………………..104 Participant Characteristics…………………………..……………………………………..104 Prenatal Serum PFAS Concentrations……...……………………………………………..105 vii Prenatal Serum PFAS and Fetal Growth Outcomes……………………………………..106 Discussion………………………………………………………………………………108 Conclusion……………………………………………………………………………...111 Supplemental Material………………………………………………………………….112 CHAPTER SIX Prenatal Ambient Air Pollution Exposure and In Utero Fetal Growth…………………………115 Abstract…………………………………………………………………………………115 Introduction……………………………………………………………………………..116 Methods…………………………………………………………………………………119 Sample………………………………………………………………………………………….119 Prenatal Ambient Air Pollution Measurements…………………………………………...120 Fetal Growth Measurements…………………………………………………………….......121 Covariates…………………………………………………………………………………...…122 Statistical Analysis…………………………………………………………………………....123 Results…………………………………………………………………………………..125 Participant Characteristics…………………………………………………………………..125 Ambient Air Pollution………………………………………………………………………...126 Fetal Growth Outcomes……………………………………………………………………...126 Prenatal Air Pollution Exposure and Fetal Growth……………………………………...127 Discussion………………………………………………………………………………130 Conclusion…………………………………..………………………………………….134 Supplemental Material………………………………………………………………….135 CHAPTER SEVEN…………………………………………………………………………….141 Summary………………………………………………………………………………..141 Conclusions and Implications…………………………………………………………..144 Future Studies…………………………………………………………………………..147 Conclusions……………………………………………………………………………..150 REFERENCES……………………………..……….………………………………………….151 viii LIST OF TABLES 3.1. Maternal Demographics (N=733)……………………………………………...…….………60 3.2. Infant Characteristics (N=733)………………………………………….………….………..61 4.1. Demographics of 342 Mother-Infant Dyads………………………...…………….………....76 4.2. Distribution of PFAS (ng/mL) Concentrations in Maternal Blood Serum (N=342)……....….77 4.3. Adjusted Single PFAS Exposure Models with Continuous Birthweight and Sex or Parity Specific BW-for-GA Z-Scores (N=342)……………….…………………………….……80 4.4. Linear Regression Model Results for 5 PFAS Compounds Simultaneously on Continuous Infant Birthweight and Sex or Parity Specific BW-for-GA Z-Scores (N=342)……...81 4.s1. Median PFAS Analyte Concentrations by Maternal Demographic Characteristics...............88 4.s2. Univariate and Adjusted Models for Independent Prenatal PFAS Compounds and Gestational Age at Birth, in Weeks (N=342)……………………………………………….…….89 4.s3. Maternal Serum PFAS and Birthweight Models Restricted to Participants Without Preeclampsia/Eclampsia, Gestational Hypertension or Gestational Diabetes Mellitus (GDM) (N=261)………………………………..…...…………………………………..90 4.s4. Maternal Serum PFAS and Birthweight Models Restricted to Samples Collected Prior to 30 Weeks Gestation (N=239)……………………………………………….…………………91 5.1. Participant Characteristics (N=335)………………………………………………………..105 5.2. Distribution of PFAS (ng/mL) Concentrations in Maternal Blood Serum (N=335)……...…106 6.1. Participant Characteristics (N=281)……………………………………………………..…125 6.2. Distributions of Pregnancy Average and Weekly Average Ambient Air Pollutants….….....126 6.3. Descriptive Statistics of Fetal Growth Outcomes………………………...…………….…..127 6.4. Linear Regression Results for Prenatal Air Pollution Exposure and Fetal Growth….…..….130 6.s1. Demographics of 863 Participants Within the MADRES Study……………………..……135 ix LIST OF FIGURES 1.1. a. Chemical Structure of Perfluorooctanoate (PFOA), and b. Perfluorooctane sulfonate (PFOS)………………………………………………………………………………………….…8 2.1. Ultrasound Scans of Abdominal Circumference (AC), Femur Length (FL), and Biparietal Diameter (BPD)……………………………………………………………………….45 4.1. Consort Diagram of Included Mother-Infant Dyads…………………………………………70 4.2. Spearman Correlation Coefficients for Natural Log Transformed PFAS (ng/mL) in Maternal Blood Serum (N = 342)………………………………………………………….……..78 4.s1. Directed Acyclic Graph (DAG) of PFAS Exposure and Infant Birthweight………………..87 5.1. Consort Diagram of Included Participants…………………………………………………...97 5.2. Prenatal Serum PFAS Concentrations and Fetal Head Circumference (N=799 scans)……..107 5.3. Prenatal Serum PFAS Concentrations and Fetal Biparietal Diameter (N=806 scans)…...…108 5.s1. Directed Acyclic Graph (DAG) of PFAS Exposure and Fetal Growth…………………….112 5.s2. Prenatal Serum PFAS Concentrations and Fetal Abdominal Circumference in mm (N=800 scans)…………………………………………………………………………………..112 5.s3. Prenatal Serum PFAS Concentrations and Fetal Femur Length in mm (N=800 scans)........113 5.s4. Prenatal Serum PFAS Concentrations and Estimated Fetal Weight in g (N=804 scans)…..113 5.s5. Sensitivity Analyses for Prenatal PFOA Exposure and Biparietal Diameter and Head Circumference ……………………………..………………………………………..…...114 6.1. Consort Diagram of Included Participants………………………………………….……....120 6.2. DLM Results Relating Weekly PM2.5 Exposure to Estimated Fetal Weight per IQR…...…..129 6.3. DLM Results Relating Weekly PM2.5 Exposure to Fetal Abdominal Circumference per IQR.........................................................................................................................................129 x 6.4. DLM Results Relating Weekly PM10 Exposure to Fetal Abdominal Circumference (mm) per IQR…...……………………………………………………………………………...130 6.s1. Spearman Correlations of Pollutants Weeks 1-32 of Gestation……………………..……..136 6.s2. Pearson Correlations of Fetal Growth Outcomes With Infant Birthweight…………..……137 6.s3. DLM Model Results for PM10 and Fetal Abdominal Circumference Stratified by Fetal Sex……………………………………………………………………………..…………138 6.s4. Results of DLM Models After Additionally Adjusting for Chronic/Gestational Diabetes and Hypertension…………………..………………………………………….……....139 6.s5. Results of DLM Models After Removing Participants Who Reported Any Personal Smoking (N=7)…………………………………………………………………………………140 xi ABSTRACT The gestational period of the life course is a particularly susceptible period to the effects of environmental exposures. Such exposures can negatively impact fetal development during pregnancy and lead to increased disease morbidity in later childhood and adulthood. Studies have documented that exposure to chemical contaminants differs by racial and ethnic groups, as does the likelihood of having an infant be born with low birthweight (<2,500 grams). Additionally, populations facing health disparities tend to live in communities that house multiple sources of pollution due to a variety of factors such as the proximity to major traffic sources and industrial facilities. Beyond higher exposure levels, these communities may be more vulnerable to health effects from these pollutants due to the compounded challenge of structural inequities related to healthcare access, economic disadvantage, and social stressors. This dissertation investigates the influence of two environmental exposures possibly experienced during pregnancy and their impacts on measures of fetal growth within the Maternal And Developmental Risks from Environmental and Social Stressors (MADRES) pregnancy cohort—a cohort of predominately low-income Hispanic participants in urban Los Angeles, California. We investigate per-and polyfluoroalkyl substances (PFAS) measured in prenatal blood serum and the association with infant birthweight and in utero fetal growth trajectories. We demonstrate that prenatal exposure to perfluorooctanoic acid (PFOA) is associated with fetal growth in two ways: 1) by reducing infant birthweight and 2) by impacting in utero fetal growth with reductions in head circumference and biparietal diameter over pregnancy. Notably, we provide novel evidence that the association of PFOA exposure on in utero fetal growth may be modified by maternal perceived stress experienced during pregnancy. This dissertation also examines how ambient air pollution experienced in the three months prior to conception, as well as across gestation, influences fetal xii growth measured in the third trimester of pregnancy within MADRES participants. This is one of the first studies to assess these associations using distributed lag modeling (DLM) techniques, which allowed critical windows of exposure to be identified with fine temporal resolution. We found that particulate matter with an aerodynamic diameter of 2.5 μm or less (PM2.5) during weeks four to sixteen of gestation was associated with significantly lower estimated fetal weight measured in the third trimester of pregnancy and PM2.5 exposure during weeks one to twenty-three weeks of pregnancy was associated with significantly decreased fetal abdominal circumference. Moreover, particulate matter with an aerodynamic diameter of 10 μm or less (PM10) exposure during weeks six to fifteen of pregnancy was associated with significantly decreased third trimester fetal abdominal circumference. Overall, we provide strong evidence that two different classes of environmental chemicals produce negative effects on fetal growth within a health disparities population. These findings have the potential to influence future policy decisions about harmful contaminants to help mitigate the future effects of exposure on fetal growth. Further research should explore mixture methods approaches for these exposures experienced concurrently and elucidate which specific components of particulate matter influence adverse fetal growth across pregnancy. 1 OVERVIEW The ‘Developmental Origins of Health and Disease (DOHaD)’ hypothesis, suggests that the early in utero environment may impact future health during childhood and adulthood [1]. The environment is generally defined as risk factors outside of genetics [2]. Environmental risk factors related to gestation include, but are not limited to, nutrition, smoking, alcohol use, socioeconomic factors, as well as elements associated with the physical environment such as air pollution and chemical exposures [3, 4]. It is common for these factors to be interdependent or to have synergistic relationships with one another. Previous literature has also shown that the environment has the ability to influence pregnancy and can increase the likelihood of adverse birth outcomes. Adverse developmental outcomes that are related to these exposures include: fetal growth restriction, functional abnormalities, birth defects, preterm delivery (birth <37 weeks gestation), low birthweight (<2500 grams), spontaneous abortion (miscarriage prior to 20 weeks gestation) and stillbirth (loss of the fetus after 20 weeks gestation) [5]. As of 2018, the United States Environmental Protection Agency (EPA) has reported that over 86,000 chemicals exist for commercial production and use across the country [6]. Environmental chemical exposures inundate daily life due to their presence in the air, food, water, and commonly used consumer products. However, due to the biological changes that occur during pregnancy, sensitivity to chemical exposures for both the mother and fetus could be increased during this period of the life course [7]. Numerous in utero exposures to chemicals and environmental toxins have been shown to induce disease later in life [8]. The developing fetus is particularly susceptible because of the ability of various chemicals to pass the placental barrier [9, 10]. Numerous chemicals have been shown to accumulate across gestation including per- and polyfluoroalkyl substances (PFAS) which are synthetic chemicals and have consistently been found in the placenta and multiple types of measured fetal tissues [11]. 2 Environmental exposures during development also have the ability to influence the placenta and the intrauterine environment, which can lead to adverse birth outcomes and ultimately impact proper fetal growth. Within the United States (US) in 2019, 8.31% of infants were born with low birthweight [12], and this rate has been slightly increasing over recent years [13]. Low birthweight has been shown to increase the likelihood of childhood obesity and subsequent metabolic health morbidities such as type 2 diabetes mellitus and cardiovascular disease later in life [14-17]. Also in 2019, 10.02% of US births were born preterm and it was the fifth straight year of increases for this rate [13]. Preterm birth has been shown to be a predictor for neurodevelopmental deficits, chronic kidney disease, reduced insulin sensitivity, hypertension, cardiovascular diseases, and chronic obstructive pulmonary diseases throughout the lifespan [18]. Health disparities exist between racial and ethnic groups as well as income levels when discussing pregnancy and birth outcomes [19]. Women with income levels below the poverty line during pregnancy have been shown to have increased adverse birth outcomes due to reduced access to care and factors related to maternal stress [20]. Within the US, Black and Hispanic populations experience higher rates of preterm birth and low birthweight compared to their non-Hispanic white counterparts [13]. Disparities in exposure levels to environmental pollutants and toxins within the US are also present, with race and income levels being significant predictors. It is well documented within the literature that racial and ethnic minorities experience disproportionate exposure to toxic air emissions due to industrial practices [21, 22] and proximity to hazardous waste facilities [23]. Previous research has also shown that communities of color within California are disproportionally burdened by environmental health hazards using the environmental justice screening tool CalEnviroScreen [24]. 3 The overall objective of my dissertation is to better understand how environmental exposures influence fetal growth, particularly in populations facing health disparities. Although pregnant mothers and developing fetuses may experience numerous environmental exposures throughout pregnancy, this dissertation only focuses on two – PFAS and ambient air pollution – and specifically assesses their influence on fetal growth within a structurally marginalized population in urban Los Angeles. Chapter 1 introduces PFAS as well as reviews the current literature on five PFAS analytes and their relationships with fetal growth measured by birthweight and growth in utero, and the influence on gestational age at birth. Chapter 2 introduces ambient air pollution and reviews the existing literature on four criteria air pollutants and their relationship with in utero fetal growth. Prior to presenting three original research studies that encompass my dissertation work, Chapter 3 provides a high-level introduction to the study design and participant characteristics of the Maternal And Developmental Risks from Environmental and Social Stressors (MADRES) study which serves as the overall pregnancy cohort in which the three studies were conducted. In my first research paper (Chapter 4), I explore the influence of prenatal PFAS exposure on infant birth outcomes including birthweight and gestational age at birth, as well as effect modification by infant sex. This study additionally defines the demographic characteristics that are associated with PFAS levels in the MADRES cohort. This study was conducted to address several gaps and potential biases from previous work. The hypothesis was that participants with higher levels of prenatal PFAS exposure would have infants with decreased birthweight. PFAS concentrations were assessed from a single blood serum sample that was collected during pregnancy. Birth outcomes were primarily abstracted from electronic medical records. Multivariate linear regression models were used to estimate the quantitative influence of prenatal 4 PFAS exposure on birth outcomes both in single pollutant and multipollutant models. This study found evidence that prenatal perfluorooctanoic acid (PFOA) exposure significantly decreased infant birthweight. Differences in PFAS levels were also identified based on participant race/ethnicity, income levels, education levels, country of birth, and parity. In my second paper (Chapter 5), I extend the hypothesis assessed in Chapter 4 that prenatal PFAS exposure influences fetal growth by using measures collected in utero. The literature that has previously described the association of prenatal PFAS exposure on fetal growth has primarily been conducted using infant birthweight or infant growth measures (i.e. birth length and head circumference at birth). Given the importance of in utero growth trajectories in predicting later life outcomes [25, 26], I wanted to specifically explore how this chemical class influenced fetal growth during gestation. In utero fetal growth was measured across pregnancy by transabdominal ultrasounds that were sourced from prenatal electronic medical records or collected from a study- measured ultrasound in the third trimester of pregnancy. This longitudinal analysis of fetal growth trajectories was evaluated using linear mixed effects models. A secondary aim was also to explore whether maternal perceived stress experienced during pregnancy modified these associations. I was interested in exploring this possible interaction because previous literature suggests a joint effect between chemicals and psychosocial stress on fetal growth [27]. This study found that PFOA exposure during pregnancy was inversely associated with fetal head circumference and biparietal diameter. Furthermore, we found stronger effects in participants with higher perceived stress levels, providing novel epidemiological evidence that maternal stress and PFAS exposure may have a synergistic effect on fetal growth. My third research paper (Chapter 6) explored the effects of ambient air pollution exposure prior to conception as well as during gestation on in utero fetal growth. Ambient air pollutants 5 were estimated at participant residences using inverse-distance-squared weighted spatial interpolation, and fetal growth was measured via a study-measured transabdominal ultrasound during the third trimester of pregnancy. To address a current gap within the literature of primarily using overall prenatal average air pollution exposure on fetal growth (i.e. average exposure from date of conception to date of ultrasound scan), we used distributed lag model (DLM) techniques to identify critical windows of exposure. Additionally, we investigated effect modification by fetal sex for these associations. This study identified several critical weeks of exposure to particulate matter during early to mid-pregnancy on fetal abdominal circumference and estimated fetal weight. In Chapter 7, I summarize the main findings of my dissertation and discuss their possible public health implications for children’s health. I conclude with suggested possible next steps for future research. 6 CHAPTER 1 Introduction to Per and Polyfluoroalkyl Substances (PFAS) There are more than 4,700 individual compounds under the umbrella term per and polyfluoroalkyl substances (PFAS) [28]. These anthropogenic organofluorine substances have been in production since the 1940’s and are known for their water and oil repelling properties due to the strong and stable bond created between carbon and fluorine (C-F) [29, 30]. These substances are considered to be long-chain (C>7), short-chain (C=4-7), or ultra-short-chain (C=2-3) PFAS, with previous research heavily focused on long-chain PFAS [31]. Shorter chain PFAS are often more persistent, however, longer chain PFAS are still remarkably stable and upon chemical decomposition frequently degrade into shorter chained PFAS [32, 33]. These substances have bio- accumulating properties and the half-lives for PFAS are several years, which is why they are often deemed “forever chemicals” [34-37]. Common household and industrial products that utilize PFAS include non-stick cookware, stain and water-resistant fabrics, metal plating, fire-fighting foams, fast food packaging, and personal care products [28, 31, 36]. The widespread use of PFAS have caused the compounds to be consistently found in the environment. PFAS chemicals mobilize by leaching into landfills, groundwater, waste runoff, and bodies of water. They are also present in dust particles and in the food web [28, 38]. The primary exposure routes to humans are through drinking water, consuming contaminated food (i.e. fish or food prepared in packaging with PFAS), and occupational exposures [39-43]. For these reasons, PFAS have been detected in virtually all samples of the nationally representative National Health and Nutrition Examination Survey (NHANES) cohort [44]. The two compounds that have typically had the highest detected levels in samples, and hence are the most researched, are perfluorooctanoic acid (PFOA) and perfluorooctanesulfonic 7 acid (PFOS). Their chemical structures are shown in Figure 1. Animal and in vitro studies have shown PFOA and PFOS to have toxic effects on the liver, kidneys, immune system, cardiovascular system, pulmonary system, and reproductive system, as well as disruption to the thyroid [45-54]. As a result of these findings, the EPA PFOA Stewardship Program in 2006 invited the eight principal companies that produced PFOA and PFOA-related chemicals to phase out production by 2015 [55]. In recent years, the Office of Environmental Health Hazard Assessment (OEHHA) has declared certain substances under the chemical class to be reproductive toxicants and are included in the Safe Drinking Water and Toxic Enforcement Act of 1985 (California’s Proposition 65) [56]. In November 2017, both PFOA and PFOS were listed for the connection to developmental toxicity, while most recently in December 2021, PFNA and its salts were listed for being related to male reproductive toxicity [56]. Internationally, only certain countries have restrictions. Examples include the prohibition of PFOA and PFOS in Canada [57] and the elimination of PFOA and PFOS and commitment to phase out all PFAS via regulating levels in water and food within the European Union [58]. While PFOA and PFOS have been phased out of manufacturing by regulations in certain countries, this is not the case globally and they can be imported into the US through consumer products [59]. Even in recent years, PFOA and PFOS have been consistently found in US groundwater and other sources of drinking water, which have been tied to industrial sites, military fire training areas, and wastewater treatment plants [49]. Additionally, other PFAS chemicals including hexafluoropropylene oxide dimer acid (GenX) and perfluorobutanesulfonic acid (PFBS) have served as replacements within the US for the substances that were phased out. There are different proposed regulations occurring at the state level, with California passing recent 8 legislation with AB-1200 to ban food packaging with PFAS chemicals and to properly disclose chemicals in cookware [60]. Several studies have analyzed the demographic characteristics related to higher PFAS exposure in humans. Women pregnant with their first child have higher levels of PFAS compared to women with more children [61-66], likely because breastfeeding serves as a probable excretion route for PFAS [63, 65]. Prior studies have also reported higher levels of PFAS (PFOA and PFHxS) in non-Hispanic white women [67], including within NHANES for PFOS, PFOA, PFNA, and PFHxS [43]. Higher education levels and annual household income have also been positively associated with PFAS levels [62, 68, 69]. Obesity and overweight (measured by body mass index or waist circumference) have been positively correlated with PFAS exposure in general populations [70]. Figure 1. a. Chemical structure of Perfluorooctanoate (PFOA), and b. Perfluorooctane sulfonate (PFOS) a. b. (Sinclair et al., 2020, Chemosphere) 9 Literature Review of PFAS on Fetal Growth and Birth Outcomes This review focuses on literature pertaining to five prominent PFAS and their influence on fetal growth and birth outcomes. A majority of studies have investigated PFAS impacts on birthweight as a surrogate measure of fetal growth; however, we also review the limited literature that has investigated PFAS associations with in utero measured fetal growth outcomes. In addition, we examine studies that investigated prenatal PFAS associations with gestational age at birth in order to assess whether gestational age may mediate the association with infant birthweight. For the purpose of this review, we limit studies to those that assessed PFAS exposure in maternal bio- specimens (i.e. maternal blood serum/plasma during pregnancy or cord blood serum/plasma at delivery) and did not include studies with PFAS concentrations measured in other media (i.e. levels detected in drinking water). The five PFAS are perfluorooctanoic acid (PFOA), perfluorooctanesulfonic acid (PFOS), perfluorohexanesulfonic acid (PFHxS), perfluorononanoic acid (PFNA), and perfluorodecanoic acid (PFDA). All of these compounds are long chain PFAS, with the exception of PFHxS, which is a short chain PFAS. These are also the compounds that will be assessed for their influence on fetal growth in this dissertation, given their frequency of detection in our study population (had at least 50% of samples detected out of the 14 analytes that were measured). We begin by discussing the possible biological mechanisms for how prenatal PFAS exposure influences fetal growth. Biological Mechanisms for how Prenatal PFAS Exposure Influences Fetal Growth Several biological mechanisms have been hypothesized for how prenatal PFAS exposure could influence fetal growth during gestation; however, there is no consensus within the literature. A leading mechanism is centered on hormone disruption caused by PFAS. Thyroid hormones are 10 crucial to the developing fetus to promote fetal growth [71]. Both PFOA and PFOS exposure have been shown to disrupt hormone signaling that occurs within the thyroid affecting homeostasis [72- 75]. Maternal hypothyroidism (levels below normal) during pregnancy has been connected to low birthweight [71]. Estrogen has also been shown to be critical in promoting proper fetal development [76]. In animal studies, PFAS have been shown to influence the expression of estrogen-responsive genes, and PFOA, PFOS, PFHxS, PFNA, and PFDA have indicated interference with the estrogen receptor in human in vitro studies [77-80]. An additional possible biological pathway is PFAS exposure disrupting placental homeostasis [81]. Previous literature has indicated that the placenta is vulnerable to environmental insults [82]. The placenta is a justifiable target of PFAS and shares common biological features as the kidney and liver, which are known organs where PFAS accumulates [82]. PFAS have been shown to be toxic to placental tissue in animal studies [83]. Placental dysfunction can reduce the nutrient supply available to the developing fetus and is associated with fetal growth restriction (FGR) [84, 85]. When assessing PFAS in fetal tissue, levels were similar to those seen in the placenta. PFAS levels were also higher in the liver and the lungs of fetuses compared to other measured tissues [86]. It has also been suggested that prenatal PFAS exposure may impact bone growth during fetal development. Experimental research in animals has shown that PFOA may influence the differentiation of osteoclasts and osteoblasts in bone cells [87, 88]. Lastly, immunotoxicity from PFAS exposure has been hypothesized as a biological pathway for influencing fetal growth as studies have shown reduced immune responses via antibody levels in animals and children from PFAS exposure [72, 89, 90]. Immunotoxicity would influence fetal growth either by pregnant mothers having increased susceptibility to infection during gestation, which would impact fetal 11 development, or by direct fetal toxicity due to adverse effects on the fetal immune system [91]. Prenatal PFOA Exposure Birthweight Prenatal exposure to PFOA in pregnant rats and mice has consistently led to reduced offspring birthweight [92-94]. A meta-analysis published in 2014 by Koustas et al. found that for each one unit increase in prenatal PFOA dose (milligrams per kilogram body weight (mg/kg) per day which equates to roughly 12,000 nanograms per millimeter (ng/mL)) was associated with a significant decrease in mean pup birthweight of -0.023 grams (95% CI = -0.029, -0.016), reflecting sufficient evidence for PFOA exposure to reduce fetal growth in animals [95]. A second meta- analysis from 2017 by Negri et al. looking at prenatal PFOA toxicology in mice and rats found similar results with statistically significant decreases in birthweight from prenatal 5 mg/kg body weight per day of PFOA exposure [96]. Although strong associations have been seen in the toxicology literature, it needs to be noted that the exposure levels used in these experimental studies are upwards of 1,000 times greater than observed human exposure levels seen within NHANES [97]. Investigations of prenatal PFOA exposure on infant birthweight in human observational studies have produced varying results. Although a consistent inverse relationship has been observed within the literature, statistical significance has only been reached in some, but not all, studies [98-102]. Comparison across studies is challenging due to different parameterization of the exposure variable in modeling approaches (log10, natural logarithm, untransformed continuous concentrations, or untransformed categorical assessment), and variability in confounders included in adjusted models. Similarly, multiple types of bio-specimens are used (maternal plasma/serum or cord blood plasma/serum), and specimen collection times vary during pregnancy across studies. 12 Importantly, the ratio between serum and plasma measurement for PFOA is 1:1 [103], but the ratio between cord and maternal plasma is 0.8 for PFOA [99, 104-107]. The serum half-life of PFOA is two to four years in humans [35] and although PFOA concentration levels across pregnancy remain highly correlated, serum levels tend to slightly decrease across gestation [87, 99]. It has also been shown that PFOA has the ability to readily cross the placental barrier [107, 108] and PFOA accumulates in the placenta across gestation [86]. There have been four systematic reviews or meta-analyses published on the association of prenatal maternal PFOA exposure and infant birthweight in the last decade [72, 96, 109, 110]. A random-effects meta-analysis conducted in 2014 (Johnson et al.) included nine studies. These nine studies had blood levels assessed between the years of 1991-2008 and represented North America (USA and Canada), Europe (Denmark, Germany, Great Britain, and Norway) as well as Asia (Taiwan, South Korea, and Japan). Measurement timing included all trimesters of pregnancy and at time of birth with median PFOA levels ranging from 1.2-5.3 ng/mL. The meta-analysis included 4,149 total births. The meta-analysis found that for each ng/mL increase in PFOA concentration from maternal blood serum/plasma or cord blood serum samples, birthweight decreased by -18.9 grams (95% CI -29.8, -7.9) [109]. They concluded that the risk of bias across studies was low and that there was “sufficient” human evidence that exposure to PFOA during gestation reduces fetal growth. They also deemed the most important confounders for adjustment in models were maternal age and gestational age at birth [109]. A systematic review published in 2015 by Bach et al. reported on the results of ten studies assessing prenatal PFOA levels between 1991-2011 on continuous birthweight [72]. Geographic representation was similar to what was seen in the Johnson et al. meta-analysis, but it also included two highly exposed populations (Darrow et al. and Wu et al.), which had median PFOA levels of 13 8.7 and 17.0 ng/mL, respectively [111, 112]. The authors of the systematic review concluded that higher PFOA levels during pregnancy were associated with lower mean birthweight in eight of the ten studies, which represented 5046 pregnancies. However, the magnitude of the association differed across studies and statistical significance was only reached in three of the studies [99, 100, 112]. The potential confounders that the authors deemed most important were parity, body mass index (BMI), and socioeconomic status [72]. The authors also concluded that effect modification by infant sex should be considered, as female infants may be more vulnerable to PFOA as identified in stratified results [72, 100]. In 2017, an updated random-effects meta-analysis (Negri et al.) was conducted from 16 articles published between 2007-2015 looking at the relationship between prenatal PFOA exposure and infant birthweight. Prenatal PFOA exposure was assessed both by untransformed values (12 studies) and log-transformed values (nine studies), although four studies provided both untransformed and transformed results and were included in both meta-analyses. Median PFOA levels ranged from 1.3 - 16.96 ng/mL and samples were measured from maternal plasma/serum during pregnancy or from cord blood at birth. Prenatal PFOA exposure was significantly associated with infant birthweight both when values were untransformed (𝛽 = -12.9g; 95% CI - 32.1, -2.38) as well as log-transformed (𝛽 = -27.1g; 95% CI -50.6, -3.6) per one unit increase in PFOA exposure [96]. There was significantly high heterogeneity across studies where values were untransformed, but low heterogeneity across studies in log transformed studies. The authors considered infant sex, gestational age at time of birth, maternal age, pre-pregnancy BMI, maternal education, parity, and prenatal smoking to be the most important potential confounders when assessing this association [96]. While no studies within the meta-analysis controlled for glomerular filtration rate (GFR), the authors noted the possible importance of GFR to be a potential 14 confounder related to both the exposure and outcome [96]. A previous study published in 2015, found that women who had a lower GFR during pregnancy had smaller babies at birth, while higher GFRs were related to lower PFAS levels due to urinary excretion routes [113]. Authors also note fish consumption may be an important confounder given PFOA is commonly found in fish [114] and also has possible positive influences on fetal growth [114]. However, no studies included fish consumption within the meta-analysis. Most recently, in 2018, Steenland et al. published an updated random effects meta-analysis on prenatal PFOA exposure and infant birthweight which included 24 studies [110]. This meta- analysis added nine new studies from the previous one by Negri et al. and also included a sensitivity analysis with a large study. This large study had previously been removed in the meta- analyses by both Johnson et al. and Negri et al. due to PFOA serum levels being estimated from age and sex based pharmacokinetic modeling with standard assumptions on water intake, body weights, and PFOA half-lives rather than measuring directly from blood serum [115]. Of the studies added, most came from Asian countries, followed by European countries, one study from the USA, and the first study conducted on the association in Oceania (Australia). Nine of the studies used PFOA levels measured in cord blood, while the remaining 15 measured levels through maternal blood (whole, plasma, and serum) during pregnancy. A total of 12,956 births were represented in the meta-analysis. Modeling techniques were used to incorporate both untransformed and logged transformed continuous estimates. In the overall analysis, the authors found that birthweight decreased by 10.5 grams (95% CI -16.7, -4.4) per ng/mL increase in PFOA in maternal or cord blood [110]. The authors reported that there was large heterogeneity across studies, which they hypothesized came from the different time points in pregnancy that the PFOA levels were assessed. Meta-analyses were stratified by those that assessed PFOA levels early in 15 pregnancy (𝛽 = -3.3 grams; 95% CI -9.6, 3.0; 5,393 pregnancies) vs. those late in pregnancy (𝛽 = -17.8 grams; 95% CI -25.0, -10.6; 7563 pregnancies) and the meta-analysis was only significant in studies later in pregnancy [110]. The author’s note that studies that had samples assessed late in pregnancy or at birth may be subject to reverse causation or residual confounding [110]. Based on the above findings, GFR may be a source of residual confounding in studies that assessed PFOA levels late in pregnancy. During pregnancy, maternal plasma volume increases throughout gestation, with a peak during 30-35 weeks [110, 116]. Verner et al. used a Monte Carlo Physiologically Based Pharmacokinetic (PBPK) model of pregnancy to decipher how much of the prenatal PFOA and birthweight association may be attributable to GFR. Models were simulated 250,000 times to represent a population and authors concluded that GFR does drive a portion of the association, but not all of it [113]. For each 1-ng/mL increase in simulated PFOA in cord plasma, birthweight decreased by 7.13 grams (95% CI -8.46, -5.80) [113]. Gestational Age at Birth It has been debated whether adjusting for gestational age at time of birth is appropriate in studies investigating infant birthweight as an outcome [117]. Bach et al. addressed this in their systematic review and stated that a way to distinguish effects on birthweight from effects on gestational age is to model gestational age at time of birth as the outcome [72]. Within this review, they concluded that four of the five studies that assessed prenatal PFOA exposure, representing 89% of pregnancies, did not find an association with gestational age at birth, and all five studies did not find an association with preterm birth (<37 weeks gestation) [100, 104, 118, 119]. Because of this, the authors determined that it was not likely for lower birthweight from PFOA exposure to be caused by shorter gestation [72]. A recent study published in 2020 also did not find prenatal PFOA exposure to be related with gestational age in the Chemicals In Our Bodies (CIOB) study 16 [120]. An additional recent study specifically tested whether gestational age mediated effects seen with first trimester PFOA concentrations on decreased infant birthweight and found no evidence [102]. In Utero Fetal Growth There have been few studies to assess PFOA on fetal growth measured in utero. In 2019, Costa et al. published the first study on maternal PFAS exposure and fetal growth assessed longitudinally by ultrasounds in 1,230 mother-child pairs within the Spanish INMA project [121]. PFOA concentrations (median=2.35 ng/mL) were measured in blood plasma samples collected at the end of the first trimester of pregnancy (mean=13.5 weeks gestation) between 2003-2008. Ultrasound scans were conducted at 12-, 20-, and 34-weeks gestation, which measured abdominal circumference (AC), biparietal diameter (BPD), femur length (FL), and estimated fetal weight (EFW). PFOA had no significant associations with fetal growth in the main analysis of this study, however, there was evidence for effect modification by maternal smoking showing negative associations for PFOA in smokers with FL −6.8% (95% CI −12.4, −1.0%) and EFW −5.7% (95% CI −11.4, 0.1%) at week 20 [121]. There was no evidence for effect modification by fetal sex. Ouidir et al. assessed the effect of prenatal PFOA levels independently on fetal growth measurements, as well as through mixture methods approaches with 11 total PFAS compounds [122]. PFOA was measured from plasma samples collected at the entry visit in 2,284 low-risk pregnant women in the National Institute of Child Health and Human Development (NICHD) Fetal Growth Studies-Singleton Cohort between 2009-2013 from clinic sites across the US. The median PFOA exposure level was 2.00 ng/mL. Significant differences across maternal race/ethnicity were observed, with the highest median levels in white participants [122]. Fetal growth was measured across weeks 16-40 of gestation with different ultrasound schedules that were randomized to 17 participants. In PFAS mixture methods, no significant associations with fetal HC, AC, or FL were observed. In individual models, PFOA was significantly positively associated with FL (exact estimates not provided, significant after FDR correction), while no associations were found for HC or AC. Models were adjusted for race/ethnicity, age, pre-pregnancy BMI, parity, education, marital status, infant sex, gestational age at ultrasound, and plasma cotinine levels [122]. Prenatal PFOS Exposure Birthweight Prenatal exposure to PFOS in pregnant rats and mice has led to reduced birthweight in offspring [74, 75, 123-126]. A recent random effects meta-analysis on the toxicological effects of PFOS in rats and mice showed significantly decreased birthweight with serum concentrations starting at 30 parts per million (ppm) in rats and 144 ppm in mice [96]. In human observational studies, where exposure concentrations are much lower [44], studies investigating prenatal PFOS exposure on infant birthweight have produced varied results, with statistical significance present only in some studies [100, 111, 118]. The PFOS serum half-life in humans is between 4-6 years [35], and the chemical has the ability to cross the placental barrier [86, 107]. Although PFOS serum levels across gestation are highly correlated, levels tend to slightly decline across pregnancy, but less so than PFOA [47, 99, 127]. However, PFOS has been shown to accumulate in the placenta across gestation [86]. Bach et al. conducted the most recently published systematic literature review on prenatal PFOS and infant birthweight [72], which was an update from the five studies reviewed by Olsen et al. in 2009 [128]. The systematic review reported results from ten studies that measured PFOS in maternal serum/plasma during pregnancy/at delivery or in cord blood at birth. The ratio between serum and plasma measurement for PFOS is 1:1 [103] and the ratio between cord and maternal 18 plasma is 0.3 for PFOS [99, 104-107]. The range of PFOS across studies ranged from <LOD (LOD=0.02 ng/mL) - 112 ng/mL and were conducted in North America, Europe, and Asia. Assessment across studies is difficult since the time period for PFOS measurement ranged across two decades, during which exposure levels changed [72], there were differing covariates included in adjusted models, and numerous exposure parameterization methods were used (untransformed, natural log, log10, or categorically). Overall, an inverse relationship was seen in the majority of studies [100, 104, 111, 118, 129], followed by estimates close to the null [99, 106, 130], and one positive association [119]. Statistical significance was only reached in three of the studies, which all had inverse associations [87, 100, 129]. The authors note that female fetuses may be more susceptible to influences from PFOS on birthweight, due to two studies finding significant inverse relationships in females [100, 129] and that effect modification by sex may be important in this association [72]. The reviewers also deemed parity, BMI, and socio-economic status to be the most important potential confounders related to both PFOS levels and infant birthweight [72]. In 2017, Negri et al. published the first random effects meta-analysis on prenatal PFOS exposure and infant birthweight consisting of 13 studies, representing nearly 8,000 pregnancies [96]. All studies measured PFOS concentrations in biological samples either during pregnancy or at birth and were conducted in North American, Asian, or European countries with varying levels of exposure levels across populations (PFOS range: <LOD – 112 ng/mL). Results were assessed by untransformed exposures and logged exposures, however, three studies overlapped between the two meta-analyses since they provided both estimates. Results of the pooled meta-analysis for untransformed PFOS (eight studies) was -0.92 grams (95% CI -3.4, 1.6) decrease in birthweight per one unit increase in prenatal PFOS, with significantly high heterogeneity across studies [96]. Within the logged pooled meta-analysis (eight studies), results concluded that for each loge ng/mL 19 increase in prenatal PFOS birthweight decreased by 46.1 grams (95% CI -80.3, -11.9) and there was low heterogeneity across studies [96]. Due to the discrepancy in significant results, the authors deemed the epidemiologic evidence for an inverse association between prenatal PFOS maternal blood concentrations and infant birthweight as “insufficient tending to moderately likely” [96]. The authors considered infant sex, gestational age at birth, maternal age, pre-pregnancy BMI, education, parity, and smoking to be the most crucial potential confounders within the association, but also that GFR and fish consumption should possibly be considered [96]. Dzierlenga et al. published an updated random effects meta-analysis in 2020 on prenatal PFOS exposure and infant birthweight consisting of 32 published manuscripts from 29 studies [131]. Studies were published between 2007-2019 (with blood samples collected since the early 1990’s) with samples sizes ranging from 85 to 3,507 participants for a total of nearly 18,000 pregnancies represented [131]. Most studies were conducted in Europe, followed by Asia, then North America, and one study from Australia. The majority of studies were longitudinal (18), and 11 studies were cross-sectional. Modeling techniques were used to incorporate both untransformed and logged transformed continuous exposure concentrations within meta-analyses. Median PFOS concentrations were lowest overall among studies conducted in Asia. The overall results of the random effects summary was −3.22 grams (95% CI −5.11, −1.33) decrease in birthweight per ng/mL increase in prenatal PFOS exposure. The authors also split the meta-analyses into studies the collected samples “early” (pre-pregnancy, first trimester, or first and second trimester combined) and “late” (second trimester, third trimester, second and third trimester combined, or cord blood). Within “early” studies (10 studies), for each ng/mL increase in prenatal PFOS, birthweight decreased by 1.35 grams (95% CI = −2.33, −0.37). For the late pregnancy group (22 studies), per ng/mL in prenatal PFOS, birthweight decreased by 7.17 grams (95% CI = -10.93, - 20 3.41) [131]. Authors deemed that due to the differences in results from blood sampling, that the association was weakly or not supportive of a casual association, but did note that the strong inverse association seen in Asian studies (where virtually all blood samples were taken at time of delivery) could not be completely explained by time of blood draw [131]. The authors did not evaluate the possible biases across individual studies, but suspected little other than the potential bias attributable to timing of the blood draw or failing to adjust for parity [131]. Gestational Age at Birth There have been conflicting findings on the association between prenatal PFOS exposure and gestational age at time of birth; however, many studies have found no association [111, 119, 120, 132-134]. There have been other studies that have shown PFOS to be inversely associated with gestational age and increased risk for pre term birth including a large birth cohort in the United States that measured PFOS early in pregnancy [135] and a Taiwanese birth cohort which measured PFOS in cord blood [118]. The influence of PFOS on gestational age is of particular interest due to the concern that gestational age may be a mediator for the effect on infant birthweight. The systematic review conducted by Bach et al. specifically addressed this and three of the four studies that assessed prenatal PFOS and gestational age found no association (992 out of 1421 pregnancies) and five of the six studies that assessed prenatal PFOS found no association between PFOS and preterm birth (9292 out of 9722 pregnancies) [72]. Therefore, the authors concluded it is not likely that the association between PFOS and infant birthweight is caused by shorter gestation [72]. An additional recent study specifically tested whether gestational age mediated effects seen between PFOS and infant birthweight and found no evidence [102]. In Utero Fetal Growth Few studies to date have looked at prenatal PFOS exposure on the effects of fetal growth 21 (cross-sectional or longitudinal assessments) by ultrasound fetal biometry measurements. Costa et al. measured prenatal PFOS concentrations from blood serum collected towards the end of the first trimester (mean=13.5±1.7 weeks gestation) in 1,230 mothers participating in the Spanish IMNA- project between 2003 and 2008 [121]. Median PFOS levels were 6.05 mg/mL for the study population. Fetal biometry measurements were collected via prenatal ultrasound at 12-, 20-, and 34-weeks gestation and included abdominal circumference (AC), femur length (FL), biparietal diameter (BPD), and estimated fetal weight (EFW). No significant associations were identified in the main analysis looking at prenatal PFOS exposure and fetal biometry outcomes at any trimester. There was also no evidence for effect modification by the sex of the fetus. The authors did find evidence that maternal smoking modified the association. Smokers made up 31% of the sample, and among smokers positive associations were observed with PFOS and BPD at the 34-week ultrasound [121]. Ouidir et al. assessed the effect of prenatal measured PFOS independently on fetal growth measurements, as well as through mixture methods approaches with 11 total PFAS compounds [122]. PFOS was measured from plasma samples collected at the entry visit in 2,284 low-risk pregnant women in the National Institute of Child Health and Human Development (NICHD) Fetal Growth Studies-Singleton Cohort between 2009-2013 from clinic sites across the US. The median PFOS exposure level was 5.16 ng/mL. Significant differences across maternal race/ethnicity were observed, with the highest median levels in white participants [122]. Fetal growth was measured across weeks 16-40 of gestation with different ultrasound schedules that were randomized to participants. In PFAS mixture methods, no significant associations with fetal HC, AC, or FL were observed. In individual models, PFOS suggested an inverse association with HC (exact estimates not provided, significant only prior to FDR correction), while no associations were observed for 22 HC or AC. Models were adjusted for race/ethnicity, age, pre-pregnancy BMI, parity, education, marital status, infant sex, gestational age at ultrasound, and plasma cotinine levels [122]. Prenatal PFHxS Exposure Birthweight The reproductive toxicology evidence for PFHxS is not as well studied as PFOA and PFOS. A study published in 2018 that was conducted in rats found no overt toxicity in dams and offspring, but PFHxS was associated with decreased birthweight in male pups [136]. Most recently, published in 2021, Tetzlaff et al. found prenatal PFHxS exposure in rats to be associated with decreased birthweight in both male and female offspring [137]. The relationship between prenatal PFHxS exposure and infant birthweight in humans has been assessed in several observational studies with varying results; however, no formal review has been conducted. PFHxS has been shown to readily cross the placental barrier [86] and the human half-life for PFHxS is 5-9 years [35, 138]. In addition, PFHxS levels have been shown to stay consistent across gestation, unlike PFOA and PFOS [87]. Three studies identified positive associations between prenatal PFHxS exposure and infant birthweight, although none reached statistical significance. Hamm et al., in 2008, published on the association of prenatal PFHxS exposure and infant birthweight in 252 women using hospital data from Alberta, Canada. Serum samples were collected in 2005 and 2006 at 15-16 weeks gestation (median PFHxS= 0.97 ng/mL). After adjusting for maternal age, race, parity, maternal weight and height, smoking, infant gender and gestational age at birth, the mothers in the highest tertile had infants weigh on average 26 grams more compared to those in the lowest tertile, although the association was not statistically significant [119]. Meng et al., in 2018, published an updated pooled analysis using the Danish National Birth Cohort consisting of 2,120 mother-infant pairs. 23 PFHxS levels were measured early in pregnancy between 1996 and 2002 and median levels were 1.0 ng/mL. After adjusting for infant sex, infant birth year, gestational week of blood draw, maternal age, parity, socio-occupational status, pre-pregnancy BMI, smoking and alcohol intake during pregnancy, per doubling of PFHxS exposure infant birthweight increased by 1.2 grams (95% CI -28.3, 30.7) [139]. Most recently, in 2021, Chen et al. published on the association of repeated prenatal PFHxS measured at all three trimesters in maternal plasma as well as in cord blood at birth and the relationship on infant birthweight within the Shanghai Birth Cohort (N=214). Blood samples were collected between 2013 and 2015 and median PFHxS levels were 0.64 ng/mL in the first trimester, 0.67 ng/mL in the second trimester, 0.67 ng/mL in the third trimester and 0.44 ng/mL from cord blood samples. PFHxS was modeled with the natural log transformation and adjusted for maternal age, BMI, education level, occupation, income, fetal sex, parity, gestational age at birth, smoking, and alcohol consumption. Non-significant positive associations were observed with prenatal PFHxS from the first, second, and third trimester with birthweight. The relationship with cord blood PFHxS levels and birthweight was inverse, but not significant [87]. Several studies have observed an inverse relationship with prenatal PFHxS exposure and infant birthweight, although only one reached statistical significance [100]. Maisonet et al. in 2012 published a study looking at prenatal PFHxS concentrations measured in maternal serum (median=30 weeks gestation) and infant birthweight in mother-daughter dyads participating in the Avon Longitudinal Study of Parents and Children (ALSPAC) study (N=422). Serum samples were collected between 1991 and 1992 and median PFHxS levels were 1.6 ng/mL. After adjusting for gestational age at birth, pre-pregnancy BMI, parity, and maternal smoking during pregnancy, infants born to mothers in the highest PFHxS tertile weighed on average 108 grams less (95% CI 24 -206, -10) than mothers in the lowest tertile [100]. Marks et al., published in 2019, investigated the same association within ALSPAC, but in mother-son dyads. Serum samples were also collected in 1991 and 1992, but median gestation at time of blood sample was 15 weeks gestation (median PFHxS=1.9 ng/mL). Within the 447 mother-son dyads, an inverse association was identified with PFHxS when ran continuously as well as in tertiles, but the association was not statistically significant. Covariates were slightly modified and included maternal age, pre-pregnancy BMI, folic acid use, smoking during pregnancy, alcohol during pregnancy, parity, and gestational age at time of blood sample [140]. Manzano-Salgado et al., published in 2017, looked at PFHxS measured in maternal plasma (mean =12.3 weeks gestation) from samples collected between 2003 and 2008 from 1,185 mothers within the Spanish INMA cohort and infant birthweight. The median PFHxS level was 0.58 ng/mL. After adjusting for maternal age, parity, pre-pregnancy BMI, and fish intake during pregnancy, prenatal PFHxS had a negative non-significant association with infant birthweight and no modification by infant sex was observed [133]. Starling et al., published in 2017, assessed prenatal PFHxS levels within 628 participants in the Healthy Start cohort study based in Colorado, USA, and infant birthweight. PFHxS levels were measured in blood serum samples taken at a median of 27 weeks gestation between 2009 and 2014. Median PFHxS levels were 0.8 ng/mL and values were transformed to the natural log and ran continuously as well as assessed in tertiles. After adjusting for maternal age, pre-pregnancy BMI, race/ethnicity, education, gestational weight gain, smoking during pregnancy, parity, gestational age at blood draw, infant sex, and gestational age at birth, for each log unit increase birthweight decreased by 13.5 grams (95% CI -50.7, 23.7). [101]. Kashino et al. published a study in 2020 looking at prenatal PFHxS measured in the third trimester of 17,869 mother-child pairs from the Japanese Hokkaido Study on Environment and Children’s Health study. Maternal plasma was collected 25 between 2003 and 2009 and the median PFHxS level was 0.3 ng/mL. PFHxS was modeled as the log10 and after adjusting for gestational age at birth, maternal age, pre-pregnancy BMI, parity, infant sex, maternal educational level, and plasma cotinine concentration during pregnancy, a non- significant inverse association was observed with infant birthweight (𝛽 = -3.0 grams; 95% CI - 60.5, 54.4) [141] . There have also been several studies that have found results close to the null, indicating no association between prenatal PFHxS and infant birthweight. Monroy et al., published in 2008, found no association with prenatal PFHxS and infant birthweight within a small subset of the Canadian Family Study (estimates were not provided) [130]. Lee et al., in 2016, published a cross- sectional study looking at prenatal PFHxS measured in cord serum samples at time of birth in 85 mothers delivering at a hospital in South Korea measured in 2008 and infant birthweight. Mean PFHxS levels were 0.60 ng/mL. After adjusting for gestational age at birth, maternal age, infant gender, and clinician, there was no relationship with PFHxS (natural log transformed) and infant birthweight (𝛽 = 0.01 grams; 95% CI -0.17, 0.21) [142]. Workman et al., published in 2019, assessed the association between prenatal PFHxS measured in the second or third trimester of pregnancy through maternal blood serum samples of 414 women between 2010 and 2011 within the Canadian Healthy Infant Longitudinal Development (CHILD) study. Median PFHxS levels were 0.44 ng/mL. Multiple linear regressions used the natural log transformation for PFHxS exposure and after adjusting for maternal age, maternal smoking during pregnancy, high blood pressure during pregnancy, parity, infant sex, and gestational age at birth a non-significant association was seen with prenatal PFHxS and birthweight (slope estimates were not provided) [143]. Wilkstrom et al., published in 2020, assessed the association between prenatal PFHxS exposure on infant birthweight within 1,533 participants of the Swedish Environmental, 26 Longitudinal, Mother and child, Asthma and allergy (SELMA) study. PFHxS was measured in maternal serum between 2007 and 2010 in the first trimester of pregnancy (median=10 weeks gestation) with the median PFHxS level being 1.23 ng/mL. After adjusting for maternal weight, parity, cotinine levels and gestational age at time of birth, no significant association was observed with PFHxS and birthweight when modeled continuously which was natural log transformed (𝛽 = -0.1 grams; 95% CI -38, 38) or when assessed by quartiles [102]. Gestational Age at Birth Fewer studies have looked at the association of PFHxS on gestational age at birth and none have found significant results. Manzano-Salgado et al., Workman et al., Meng et al., and Hamm et al., discussed above, all found no association between prenatal PFHxS exposure and gestational age at birth [119, 133, 139, 143]. Sagiv et al., published in 2018, assessed prenatal PFHxS exposure and gestational age at birth and risk of preterm birth. Maternal blood plasma was collected between 1999 and 2002 at a median of nine weeks gestation within 1,645 mothers in the Project Viva study conducted in Massachusetts, USA. Median PFHxS levels were 2.4 ng/mL and associations were assessed both continuously and in quartiles. After adjusting for maternal age, race/ethnicity, education, prenatal smoking, parity, history of breastfeeding, pre-pregnancy BMI, paternal education, household income, child sex, and gestational age at blood draw, per unit increase in PFHxS gestational age at time of birth increased by 0.02 weeks (95% CI -0.04, 0.07). In addition, there was no significant increase in odds of preterm birth with PFHxS exposure (OR=1.2; 95% CI 1.0, 1.4) [135]. Eick et al., published in 2020, assessed the association between prenatal PFHxS measured in maternal whole blood between 12- and 28-weeks gestation and gestational age at birth within the Chemicals In Our Bodies (CIOB) study conducted in San Francisco, CA, USA. Samples were collected between 2014 and 2018 and the median PFHxS level was 0.33 ng/mL. PFHxS was 27 modeled in tertiles and after adjusting for maternal age, maternal race/ethnicity, pre-pregnancy BMI, maternal education, smoking status, parity, and food insecurity there was no association with gestational age at birth [120]. Gardener et al., published in 2021, assessed prenatal PFHxS exposure and gestational age at birth and risk of preterm birth in 433 women enrolled in the Vanguard Pilot Study of the National Children’s Study. Serum samples were mostly collected in the third trimester of pregnancy and median PFHxS levels were 0.5 ng/mL. PFHxS was modeled in quartiles and after adjusting for age, education, race, pre-pregnancy BMI, prenatal smoking, parity, and gestational age at time of blood sample, there was no association with gestational age or odds of preterm birth [144]. Wilkstrom et al. also did a formal mediation analysis of gestational age on the association of prenatal PFHxS exposure on infant birthweight and found no evidence [102]. In Utero Fetal Growth There have only been two studies to date that has looked at the effect of prenatal PFHxS exposure directly on fetal growth measurements [121]. In 2019, Costa et al. assessed PFHxS levels measured at the end of the first trimester (mean=13.5±1.7 weeks gestation) between 2003 and 2008 and fetal biometry measurements within 1,230 mothers participating in the Spanish IMNA- project. The median PFHxS level was 0.58 ng/mL. Fetal biometry measurements were collected via prenatal ultrasound at 12-, 20-, and 34-weeks gestation and included abdominal circumference (AC), femur length (FL), biparietal diameter (BPD), and estimated fetal weight (EFW). No significant associations were identified within the main analysis looking at prenatal PFHxS exposure and fetal biometry outcomes at any trimester. There was also no evidence for effect modification by the sex of the fetus. The authors did find evidence that maternal smoking modified 28 the association for BPD. Smokers made up 31% of the sample, and among smokers, positive associations were observed with PFHxS and BPD at the 34-week ultrasound [121]. Ouidir et al. assessed the effect of prenatal measured PFHxS independently on fetal growth measurements and through mixture methods approaches with 11 total PFAS compounds [122]. PFHxS was measured from plasma samples collected at the entry visit in 2,284 low-risk pregnant women in the National Institute of Child Health and Human Development (NICHD) Fetal Growth Studies-Singleton Cohort between 2009-2013 from clinic sites across the US. The median PFHxS exposure level was 0.71 ng/mL. Significant differences across maternal race/ethnicity were observed, with the highest median levels in white participants [122]. Fetal growth was measured across weeks 16-40 of gestation with different ultrasound schedules that were randomized to participants. In PFAS mixture methods, no significant associations with fetal HC, AC, or FL were observed. In individual models, PFHxS suggested a positive association with HC (exact estimates not provided, significant after FDR correction), and a positive influence with FL (exact estimates not provided, significant after FDR correction). No association was observed with AC. Models were adjusted for race/ethnicity, age, pre-pregnancy BMI, parity, education, marital status, infant sex, gestational age at ultrasound, and plasma cotinine levels [122]. Prenatal PFNA Exposure Birthweight Few studies have looked at the toxicological effects of prenatal PFNA exposure and offspring birth outcomes. An experimental study published in 2010 conducted in mice found that prenatal PFNA exposure reduced the number of live pups at birth and survival of offspring to weaning, but saw no significant reduction in offspring birthweight [145]. An experimental study in rats found that pups exposed to 5mg/kg PFNA prenatally had significantly lower birthweight 29 than controls [146] and a recent study in mice concluded that the developmental toxicity of PFNA in mice is comparable to that of PFOS and PFOA [147]. Human research via observational studies on the association of prenatal PFNA exposure and infant birth outcomes have provided mixed results and no formal review of the literature has been conducted. The half-life for PFNA in humans is 3-4 years [138]. Although maternal levels of PFNA tend to marginally decrease across pregnancy [87], PFNA has been shown to accumulate in the placenta across gestation [86]. The majority of studies have indicated inverse relationships with prenatal PFNA exposure and infant birthweight, although only some have reached statistical significance [87, 102, 139- 141]. Starling et al., published in 2017, assessed prenatal PFNA levels within 628 participants of the Healthy Start cohort study based in Colorado, USA, and infant birthweight. PFNA levels were measured in blood serum samples taken at a median of 27 weeks gestation between 2009 and 2014. Median PFNA levels were 0.4 ng/mL and values were transformed to the natural log as well as assessed dichotomously at the median in multiple linear regression models. After adjusting for maternal age, pre-pregnancy BMI, race/ethnicity, education, gestational weight gain, smoking during pregnancy, parity, gestational age at blood draw, infant sex, and gestational age at birth, for each log unit increase of PFNA, birthweight decreased by 57.6 grams (95% CI -104.1, -11.2). Significant associations were also seen in dichotomous analyses (𝛽 = -92.1 grams; 95% CI -150.6, -33.6). Additionally, it was assessed whether maternal glucose levels mediated the association between prenatal PFNA and birthweight, and found that 2.5% (percent mediated calculated as natural indirect effect divided by total effect × 100%) of the association was mediated by gestational glucose [101]. Meng et al., in 2018, published an updated pooled analysis using the Danish National Birth Cohort consisting of 2,120 mother-infant pairs. PFNA levels were measured early in pregnancy between 1996 and 2002 and median levels were 0.5 ng/mL. After adjusting for 30 infant sex, infant birth year, gestational week of blood draw, maternal age, parity, socio- occupational status, pre-pregnancy BMI, smoking and alcohol intake during pregnancy, per doubling of PFNA exposure infant birthweight decreased by 36.3 grams (95% CI -70.6, -2.0). When assessed categorically, the highest quartile of maternal prenatal PFNA exposure had infants that weighed on average 81.2 grams less (95% CI -147.1, -15.4) compared to maternal PFNA exposure in the first quartile [139]. Marks et al., published in 2019, investigated prenatal PFNA exposure and infant birthweight in 447 mother-son dyads participating in the Avon Longitudinal Study of Parents and Children (ALSPAC) study. Serum samples were collected in 1991 and 1992 with a median gestation at time of blood sample of 15 weeks (median PFNA=0.4 ng/mL). After adjusting for maternal age, pre-pregnancy BMI, folic acid use, smoking during pregnancy, alcohol during pregnancy, parity, and gestational age at time of blood sample mothers in the highest tertile had sons weigh 133 grams (95% CI -265.41, -0.64) less at birth than mothers in the first tertile. Strong inverse effects were also seen with continuous PFNA (𝛽 = -169.58 grams; 95% CI -448.31, 109.16), but results were not significant [140]. Kashino et al. published a study in 2020 looking at prenatal PFNA measured in the third trimester of 17,869 mother-child pairs from the Japanese Hokkaido Study on Environment and Children’s Health study. Maternal plasma was collected between 2003 and 2009 and the median PFNA levels were 1.2 ng/mL. PFNA was modeled as the log10 and after adjusting for gestational age, maternal age, pre-pregnancy BMI, parity, infant sex, maternal educational level, and plasma cotinine concentration during pregnancy, a significant inverse association was observed with infant birthweight (𝛽 = -96.2 grams; 95% CI -165.3, -27.1) [141]. Wilkstrom et al., published in 2020, assessed the association between prenatal PFNA exposure on infant birthweight within 1,533 participants of the Swedish Environmental, Longitudinal, Mother and child, Asthma and allergy (SELMA) study. PFNA was measured in 31 maternal serum between 2007 and 2010 in the first trimester of pregnancy (median=10 weeks gestation) with the median PFNA level being 0.53 ng/mL. After adjusting for maternal weight, parity, cotinine levels and gestational age at birth, a significant inverse association was observed with PFNA and birthweight when modeled continuously as a per natural log unit increase (𝛽 = - 46 grams; 95% CI -89, -4). When models were stratified by sex, PFNA was only significantly inversely related to birthweight in female infants [102]. Inverse associations that did not reach statistical significance will now be reviewed, starting with most the recent results. Chen et al. published in 2021, assessed the association of repeated prenatal PFNA measured at all three trimesters in maternal plasma as well as in cord blood at birth and the relationship on infant birthweight within the Shanghai Birth Cohort (N=214). Blood samples were collected between 2013 and 2015 and median PFNA levels were 2.33 ng/mL in the first trimester, 1.36 ng/mL in the second trimester, 1.24 ng/mL in the third trimester and 0.62 ng/mL from cord blood samples. PFNA was modeled with the natural log transformation and adjusted for maternal age, BMI, education level, occupation, income, fetal sex, parity, gestational age, smoking, and alcohol consumption. Non-significant inverse associations were observed with prenatal PFNA from the first, second, third trimester, and at delivery with birthweight [87]. Manzano-Salgado et al., published in 2017, looked at PFNA measured in maternal plasma (mean=12.3 weeks gestation) from samples collected between 2003 and 2008 from 1,185 mothers within the Spanish INMA cohort and infant birthweight. The median PFNA levels were 0.66 ng/mL. After adjusting for maternal age, parity, pre-pregnancy BMI, and fish intake during pregnancy, prenatal PFNA had a negative non-significant association with infant birthweight (𝛽 = -10.27 grams; 95% CI -38.14, 17.61) and no modification by infant sex was observed [133]. Lee et al., in 2016, published a cross-sectional study looking at prenatal PFNA measured in cord serum 32 samples at time of delivery in 85 mothers delivering at a hospital in South Korea and infant birthweight in 2008. Mean PFNA levels were 0.36 ng/mL. After adjusting for gestational age at birth, maternal age, infant gender, and clinician, there was a weak inverse association per natural log increase with infant birthweight (𝛽 = -0.14 grams; 95% CI -0.39, 0.10) [142]. Workman et al., published in 2019, assessed the association between prenatal PFNA collected in the second or third trimester of pregnancy and measured in blood serum samples of 414 women between 2010 and 2011 within the Canadian Healthy Infant Longitudinal Development (CHILD) study. Median PFNA levels were 0.37 ng/mL. Multiple linear regression models used the natural log transformation for PFNA exposure and after adjusting for maternal age, maternal smoking during pregnancy, high blood pressure during pregnancy, parity, infant sex and gestational age at birth, a negative non-significant association was seen with prenatal PFNA and birthweight (slope estimates were not provided) [143]. Lastly, Monroy et al., published in 2008, found no association with prenatal PFNA exposure and infant birthweight in a Canadian cohort, although estimates of multiple linear regression models were not provided [130]. There has been one study that identified a positive, non-significant association between prenatal PFNA and infant birthweight. Chen et al., published in 2012, assessed PFNA collected in cord blood (mean PFNA level= 2.36 ng/mL) and infant birthweight in 429 mother-infant pairs participating in the Taiwan Birth Panel Study (TBPS). After adjusting for maternal age, pre- pregnancy BMI, education level, cord blood cotinine levels, type of delivery, parity, infant sex, and gestational age at birth, for each log unit increase in PFNA birthweight increased by 6.07 grams (95% CI -16.6, 28.7), which was not statistically significant [118]. 33 Gestational Age at Birth Fewer studies have looked at the association of PFNA on gestational age at birth and only one study has reported significant results [139]. Meng et al., described above, found that per doubling of prenatal PFNA exposure gestational age decreased by 1 week (95% CI -1.7, -0.3). No significant results were seen when PFNA was measured in quartiles with gestational age [139]. Manzano-Salgado et al. discussed above, found no association between prenatal PFNA exposure and gestational age at birth (𝛽 = -0.00 weeks; 95% CI -0.11, 0.11) [133]. Chen et al. 2012, discussed above, also found no association with PFNA and risk of preterm birth (OR= 0.88; 95% CI 0.71, 1.11) or with gestational age at birth (𝛽 = 0.04; 95% CI -0.06, 0.14) [118]. Workman et al., discussed above, did not find a significant association between prenatal PFNA and gestational age at birth (estimates were not provided) [143]. Sagiv et al., published in 2018, assessed prenatal PFNA exposure and gestational age at birth and risk of preterm birth. Maternal blood plasma was collected between 1999 and 2002 at a median of nine weeks gestation within 1,645 mothers in the Project Viva study conducted in Massachusetts, USA. Median PFNA levels were 0.7 ng/mL and associations were assessed both continuously and in quartiles. After adjusting for maternal age, race/ethnicity, education, prenatal smoking, parity, history of breastfeeding, pre-pregnancy BMI, paternal education, household income, child sex, and gestational age at blood draw per unit increase in PFNA gestational age at time of birth decreased by -0.07 weeks (95% CI -0.17, 0.02). Results were also not significant when assessed in quartiles. There was no significant increase in odds of preterm birth when PFNA was assessed continuously or in quartiles [135]. Eick et al., published in 2020, assessed the association between prenatal PFNA measured in maternal whole blood between 12- and 28-weeks gestation and gestational age at birth within the Chemicals In Our Bodies (CIOB) study. Samples were collected between 2014 and 2018 and the median PFNA 34 levels were 0.30 ng/mL. PFNA was modeled in tertiles and after adjusting for maternal age, maternal race/ethnicity, pre-pregnancy BMI, maternal education, smoking status, parity, and food insecurity there was no association with gestational age at birth [120]. Gardener et al., published in 2021, assessed prenatal PFNA exposure and gestational age at birth and risk of preterm birth in 433 women enrolled in the Vanguard Pilot Study of the National Children’s Study. Serum samples were mostly collected in the third trimester of pregnancy and median PFNA levels were 0.7 ng/mL. PFNA was modeled in quartiles and after adjusting for age, education, race, pre-pregnancy BMI, prenatal smoking, parity, and gestational age at time of blood sample there was no association with gestational age at birth. There was a significant quartile p for trend for PFNA and odds of preterm birth (p=0.04), but all estimates in reference to quartile one contained the null value [144]. Wilkstrom et al. also did a formal mediation analysis of gestational age on the association of prenatal PFNA exposure on infant birthweight and found no evidence [102]. In Utero Fetal Growth There have only been two studies that have looked at the influence of prenatal PFNA exposure on fetal growth measurements during pregnancy [121]. In 2019, Costa et al. assessed PFNA levels measured at the end of the first trimester (mean= 13.5±1.7 weeks gestation) between 2003 and 2008 and fetal biometry measurements within 1,230 mothers participating in the Spanish IMNA-project. The median PFNA levels were 0.65 ng/mL. Fetal biometry measurements were collected via prenatal ultrasound at 12-, 20-, and 34-weeks gestation and included abdominal circumference (AC), femur length (FL), biparietal diameter (BPD), and estimated fetal weight (EFW). No significant associations were identified within the main analysis looking at prenatal PFNA exposure and fetal biometry outcomes at any trimester. There was also no evidence for effect modification by the sex of the fetus. The authors did find evidence that maternal smoking 35 modified the associations. Smokers made up 31% of the sample, and among smokers negative associations were observed with PFNA and FL as well as PFNA and EFW at the 20-week ultrasound [121]. Ouidir et al. assessed the effect of prenatal measured PFNA independently on fetal growth measurements, as well as through mixture methods approaches with 11 total PFAS compounds [122]. PFNA was measured from plasma samples collected at the entry visit in 2,284 low-risk pregnant women in the National Institute of Child Health and Human Development (NICHD) Fetal Growth Studies-Singleton Cohort between 2009-2013 from clinic sites across the US. The median PFHxS exposure level was 0.77 ng/mL. Significant differences across maternal race/ethnicity were observed, with the highest median levels being identical in white and Asian participants, and lower in Black and Hispanic participants [122]. Fetal growth was measured across weeks 16-40 of gestation with different ultrasound schedules that were randomized to participants. In PFAS mixture methods, no significant associations with fetal HC, AC, or FL were observed. In individual models, PFNA suggested a positive association with AC (exact estimates not provided, only significant before FDR correction), and a positive influence with FL (exact estimates not provided, only significant prior to FDR correction). No association was observed with HC. Models were adjusted for race/ethnicity, age, pre-pregnancy BMI, parity, education, marital status, infant sex, gestational age at ultrasound, and plasma cotinine levels [122]. Prenatal PFDA Exposure Birthweight Minimal experimental research has been conducted to assess developmental toxicity of PFDA and the influence of prenatal PFDA exposure on offspring birthweight. An experimental 36 study published in 1989 conducted in mice found no reductions in birthweight [148]. In recent years, more studies have begun to look at prenatal PFDA exposure on infant birthweight; however, it remains the PFAS with the least literature on out of those discussed within this paper. The half- life of PFDA in humans has not yet been fully established, however, in rats it ranges between 40- 80 days for male rats and 59-75 days for female rats, which is longer than half-lives seen in other PFAS in rats [138]. PFDA has the ability to cross the placenta and maternal levels tend to slightly decrease across gestation [87]. Six studies have assessed the association between prenatal PFDA exposure and continuous birthweight. The majority of studies have identified inverse associations, although only some have reached statistical significance [102, 141]. Kashino et al. published a study in 2020 looking at prenatal PFDA measured in the third trimester of 17,869 mother-child pairs from the Japanese Hokkaido Study on Environment and Children’s Health study. Maternal plasma was collected between 2003 and 2009 and the median PFDA levels were 0.5 ng/mL. PFDA was modeled as the log10 and after adjusting for gestational age, maternal age, pre-pregnancy BMI, parity, infant sex, maternal educational level, and plasma cotinine concentration during pregnancy a significant inverse association was observed with infant birthweight (𝛽 = -138.1 grams; 95% CI -138.1, -6.3) [141]. Wilkstrom et al., published in 2020, assessed the association between prenatal PFDA exposure on infant birthweight within 1,533 participants of the Swedish Environmental, Longitudinal, Mother and child, Asthma and allergy (SELMA) study. PFDA was measured in maternal serum between 2007 and 2010 in the first trimester of pregnancy (median=10 weeks gestation) with the median PFDA level being 0.26 ng/mL. After adjusting for maternal weight, parity, cotinine levels and gestational at time of birth, a significant inverse association was observed with PFDA and birthweight when modeled continuously as a per natural log unit increase 37 (𝛽 = -58 grams; 95% CI -103, -13). In addition, mothers who were in the highest quartile of PFDA levels had infants weigh on average 69 grams less (95% CI -132, -5) compared to mothers in the lowest tertile. When models were stratified by sex, PFDA was only significantly inversely related to birthweight in female infants [102]. Three additional studies observed inverse relationships, although they were not statistically significant. Lee et al., in 2016, published a cross-sectional study looking at prenatal PFDA measured in cord serum samples in 2008 at time of delivery in 85 mothers delivering at a hospital in South Korea and infant birthweight. Median PFDA levels were 0.14 ng/mL. After adjusting for gestational age at birth, maternal age, infant gender, and clinician there was a non-significant inverse relationship with PFDA (natural log transformed) and infant birthweight (𝛽 = -0.18 grams; 95% CI -0.54, 0.17) [142]. Meng et al., in 2018, published an updated pooled analysis using the Danish National Birth Cohort consisting of 2,120 mother-infant pairs. PFDA levels were measured early in pregnancy between 1996 and 2002 and median levels were 0.2 ng/mL. After adjusting for infant sex, infant birth year, gestational week of blood draw, maternal age, parity, socio- occupational status, pre-pregnancy BMI, smoking and alcohol intake during pregnancy, per doubling of PFDA exposure infant birthweight decreased by 9 grams (95% CI -43.2, 25.2) [139]. Workman et al., published in 2019, assessed the association between prenatal PFDA collected in the second or third trimester of pregnancy and measured in blood serum samples of 414 women between 2010 and 2011 within the Canadian Healthy Infant Longitudinal Development (CHILD) study. Median PFDA levels were 0.13 ng/mL. Multiple linear regression models used the natural log transformation for PFDA exposure and after adjusting for maternal age, maternal smoking during pregnancy, high blood pressure during pregnancy, parity, infant sex and gestational age at 38 birth, a negative non-significant association was seen with prenatal PFDA and birthweight (slope estimates were not provided) [143]. One positive association with prenatal PFDA exposure and infant birthweight has been reported, although results were not significant [87]. Chen et al., published in 2021, reported on the association of repeated prenatal PFDA measured at all three trimesters in maternal plasma as well as in cord blood at birth and the relationship on infant birthweight within the Shanghai Birth Cohort (N=214). Blood samples were collected between 2013 and 2015 and median PFDA levels were 1.73 ng/mL in the first trimester, 1.56 ng/mL in the second trimester, 1.13 ng/mL in the third trimester and 0.40 ng/mL from cord blood samples. PFDA was modeled with the natural log transformation and adjusted for maternal age, BMI, education level, occupation, income, fetal sex, parity, gestational age, smoking, and alcohol consumption. Non-significant positive associations were observed with prenatal PFDA from the first, second, and third trimester with birthweight. The relationship with cord blood PFDA levels and birthweight was inverse, but also not significant [87]. Gestational Age at Birth Fewer studies have assessed prenatal PFDA exposure on gestational age at time of birth, and none have found significant associations [139, 143, 144]. Meng et al. and Workman et al., both discussed above, found no significant association with prenatal PFDA and gestational age at birth [139, 143]. Gardener et al., published in 2021, assessed prenatal PFDA exposure and gestational age at birth and risk of preterm birth in 433 women enrolled in the Vanguard Pilot Study of the National Children’s Study. Serum samples were mostly collected in the third trimester of pregnancy and median PFDA levels were 0.2 ng/mL. PFDA was modeled in quartiles and after 39 adjusting for age, education, race, pre-pregnancy BMI, prenatal smoking, parity, and GA at time of blood sample there was no association with gestational age or odds of preterm birth [144]. In Utero Fetal Growth Only one study has looked at the association of prenatal PFDA on longitudinal fetal growth via prenatal ultrasound measurements. In the Spanish INMA-Project, which has been discussed in above sections, PFDA was not included as a PFAS compound for analysis. Ouidir et al. assessed the effect of prenatal measured PFDA independently on fetal growth measurements, as well as through mixture methods approaches with 11 total PFAS compounds [122]. PFDA was measured from plasma samples collected at the entry visit in 2,284 low-risk pregnant women in the National Institute of Child Health and Human Development (NICHD) Fetal Growth Studies-Singleton Cohort between 2009-2013 from clinic sites across the US. The median PFDA exposure level was 0.25 ng/mL. Significant differences across maternal race/ethnicity were observed, with the highest median levels in Asian participants [122]. Fetal growth was measured across weeks 16-40 of gestation with different ultrasound schedules that were randomized to participants. In PFAS mixture methods, no significant associations with fetal HC, AC, or FL were observed. In individual models, PFDA suggested a positive association with HC, AC, and FL (exact estimates not provided, significant only prior to FDR correction). Models were adjusted for race/ethnicity, age, pre-pregnancy BMI, parity, education, marital status, infant sex, gestational age at ultrasound, and plasma cotinine levels [122]. Gaps Within the Literature After reviewing the current literature on prenatal PFAS exposure and fetal growth, several gaps have been identified that future studies should consider. The most palpable theme identified is the lack of studies that have directly measured fetal growth through anatomical measurements 40 of the fetus across gestation. To date, there have only been two studies [121, 122], and only one of them has included PFDA as a measured compound. Additional studies should include compounds other than PFOA and PFOS, which have dominated the current literature. Only studies published more recently have started including analyses covering a broader scope of PFAS exposures. However, this is also related to the premise that many of the most recently published manuscripts are using blood serum samples that were collected in the first decade of the 2000s [102, 133, 139, 141], which are unable to assess how legacy PFAS and PFAS replacements are influencing fetal growth at current day levels. Future studies should also consider how these highly correlated PFAS influence infant birthweight through multipollutant models and mixture models to consider the relationships these compounds have together. Only recently have studies incorporated Bayesian kernel machine regression (BKMR) modeling techniques, which have included mixtures of PFAS as well as other environmental toxins (i.e. organochlorine compounds (OCs), metals, phthalates, pesticides) [149, 150]. It is also important for future studies to consider how gestational age at time of blood sample influences results by including it as a covariate as well as considering how samples taken later in pregnancy may impact results. Another gap within the literature is the underrepresentation of people from racial and ethnic minority populations, particularly Hispanic women within the United States. It has been suggested that PFAS levels differ across race/ethnic groups, typically with levels lower in Hispanic populations compared to their non-Hispanic white counterparts within the United States [44], but the association of prenatal PFAS exposure on fetal growth is under-researched within this group. To date, no studies have looked at prenatal PFAS exposure and infant birthweight or fetal biometry measurements in a predominately Hispanic cohort, or in a cohort that represents a health disparities 41 population. Many studies looking at this association have been conducted in Asian and European countries with homogenous study populations and reflect more affluent socio-economic demographics. Of the studies that have been conducted in North America, only a few have included Hispanic women. Eick et al., Starling et al., Ouidir et al., and Sagiv et al., included Hispanic women, however, the percentage represented by this group made up only 34%, 23%, 28%, and 7% of the sample, respectively [101, 120, 122, 135]. With the Hispanic population in the United States continuing to grow [151], and health disparities being associated with race/ethnicity in the United States [152], it is important to understand how prenatal PFAS exposure influences fetal growth within this group. Lastly, additional assessment of effect modification needs to be considered within these associations. Few studies have looked at differences by fetal/infant sex, and further work is warranted. Outside of sex-specific differences, the only other effect modifiers to be explored in select studies include maternal smoking [121] and difference by race/ethnicity [122]. Research has suggested that multiple exposures to chemicals and psychosocial stress may exhibit a cumulative effect on fetal growth [153]. A current gap within the literature is that no studies to date have explored whether stress experienced during pregnancy modifies how prenatal PFAS exposure influences fetal growth. This is particularly important since literature has indicated that PFAS exposure impacts corticotrophin-releasing hormone and cortisone levels in pregnancy [154, 155] suggesting that stress may be an important pathway for PFAS exposure effects. 42 CHAPTER TWO Introduction to Ambient Air Pollution Ambient air pollution consists of a diverse mixture of gases and particles. A large number of sources of air pollution are human-made and comprise of mobile sources, which encompasses many modes of transportation including cars, trucks, and airplanes, and stationary sources such as factories, power plants, and refineries [156]. In addition, there are natural sources of ambient air pollution like wild fires, trees, dust, and sea or ocean aerosols [157]. According to the World Health Organization (WHO), 9 out of 10 people across the globe are exposed to levels of outdoor air pollution greater than advised international limits and the highest levels are seen in lower- or middle-income countries. It has been estimated that ambient air pollution contributes to over four million deaths annually [158]. In the United States, the Clean Air Act of 1971 designated criteria air pollutants that are ubiquitous and related to negative outcomes in both human health and in the environment and established primary and secondary air standards for them. Criteria air pollutants include ground-level ozone (O3), particulate matter (PM), nitrogen dioxide (NO2), carbon monoxide (CO), lead (Pb), and sulfur dioxide (SO2) [159]. For the purpose of this dissertation, the pollutants of interest are PM, O3, and NO2 as these tend to occur at elevated outdoor concentrations in the region of Southern California. Although declining trends have been observed over recent decades, additional reductions are still required to achieve health protective levels [160]. PM smaller than 10 and 2.5 micrometers (μm) in aerodynamic diameter (PM10 and PM2.5), respectively, is emitted from stationary, mobile, and natural sources, with larger particles emitted mostly from mechanical abrasion or natural dust or marine aerosol sources and smaller particles emitted or formed by combustion processes or atmospheric transformations and chemical reactions of of gaseous pollutants and aerosols [159]. PM is composed of a complex mixture of chemicals 43 and can have widely variable physical and chemical properties and health risks based on its sources and properties [161]. PM2.5 has the ability to penetrate deeply into the respiratory system and reach the alveolar region of the lungs [162]. Studies have shown that both PM10 and PM2.5 are associated with cardiovascular disease and premature mortality [159, 163-166]. A recent meta-analysis found prenatal exposure to PM2.5 significantly increases the odds of both preterm labor and low birthweight [167-169]. O3 is a secondary pollutant that is formed through the photochemical reaction of pollutants including nitrogen oxides and volatile organic compounds (VOCs) that are released from mobile and stationary sources in the presence of sunlight [159, 170]. O3 has been associated with numerous respiratory health risks including decreased lung function and asthma exacerbation [171-173]. Additionally, prenatal O3 exposure has also been associated with preterm birth risk [174, 175] and additional adverse birth outcomes including intrauterine growth restriction [176]. NO2 is a highly reactive gas that is primarily released into the air through the burning of fuels at high temperatures, such as from vehicles or power plants [177]. NO2 has been shown to generate reactive nitrogen-derived free radicals that have the ability to damage lung tissue [178, 179], which has been shown to significantly increase the risk of childhood asthma in a recent meta- analysis [180]. NO2 exposure is also associated with general and cardiovascular specific mortality [181]. NO2 exposure during pregnancy has been shown to have strong associations with negative birth outcomes including low birthweight, preterm birth, and infants born small for gestational age [169, 182, 183]. The concentrations of ambient air pollutants introduced above have greatly decreased within the US across recent decades [184], however, there are still large exposure and health disparities present for which groups are more likely to be exposed to higher air pollution levels 44 and have higher susceptibility to its adverse effects. It is well documented within the literature that racial and ethnic minorities experience disproportionate exposure to toxic air emissions due to nearby industrial facilities and practices in their neighborhoods and communities [185, 186]. Additionally, previous research has shown that communities of color within California are disproportionally burdened by environmental health hazards using the environmental justice screening tool CalEnviroScreen [187], and are more likely to live near traffic sources [188] compared to their non-Hispanic white counterparts. Ambient Air Pollution and Ultrasound Measures of Fetal Growth As discussed above, there is substantial evidence that exposure to ambient air pollution during pregnancy influences negative birth outcomes including low birthweight and preterm birth [167-169, 174-176, 182, 189]. However, fewer studies have accessed how ambient air pollution influences fetal growth and development in utero through measurements taken prior to birth. This literature review will focus on five distinct fetal growth parameters that are important indicators of fetal development in utero measured by prenatal ultrasounds [189, 190]. These include head circumference (HC) in millimeters (mm), which is the length along the skull bone [191], biparietal diameter (BPD) in mm which is the maximum diameter of a transverse section the fetal skull from the proximal parietal bone to the inner edge of the distal parietal bone [189, 192], femur length [193] in mm which is the length of the thigh bone [191], abdominal circumference (AC) in mm which is measured by the “ellipse facility of ultrasound equipment with the stomach bubble and a short segment of the umbilical vein at the level of the portal sinus visible” [194] and estimated fetal weight (EFW) in grams using the formula from Hadlock et al., which takes into account the above measurements [195]. Ultrasound scan measurements for AC, FL, and BPD are shown in 45 Figure 2 for reference. For the purpose of this review, trimesters are considered the following: First: 1 to <14 weeks gestation, Second: 14 to <28 weeks gestation, and Third: ≥28 weeks gestation. The association of the air pollutants NO2, PM10, PM2.5, and O3 on these fetal growth outcomes will be broken down by pollutant. The review will begin by discussing possible biological mechanisms for how ambient air pollution influences fetal growth. Figure 2. Ultrasound Scans of Abdominal Circumference (AC), Femur Length (FL), and Biparietal Diameter (BPD) Žaliūnas et al. 2018. Medicina Biological Mechanisms- Prenatal Ambient Air Pollution and Fetal Growth One of the central hypotheses for how prenatal air pollution exposure influences fetal growth is related to impacts to the placenta. The placenta has been shown to be a susceptible target of environmental insults [196]. Black carbon, which makes up a significant portion of PM mass, has been found in the fetal side of the placenta [197], and research has shown that by placental transfer, PM from the mother migrates to the lungs of the growing fetus [198]. Maternal PM exposure is additionally associated with oxidative stress in the mother leading to placental inflammation and eventually impaired transplacental oxygen and nutrient delivery to the fetus by triggered hemodynamic responses, ultimately impacting fetal growth [199]. A recent review of the literature also indicated that prenatal air pollution exposure is associated with both nitrosative stress and epigenetic changes in the placenta [200]. Additional proposed biological pathways 46 include the formation of cell-damaging lipid peroxides and reduced maternal antioxidant reserves, which in turn can have toxic effects on the growth and development of the fetus and has shown to be related to prenatal NO2 exposure [201, 202]. Prenatal NO2 Exposure and Fetal Growth Nine studies to date have assessed prenatal outdoor NO2 exposure on FL measured by ultrasound scans [203-211]; three of these studies were within subsets of the Spanish INMA Project [203, 206, 212]. Within these studies, three have found significant inverse associations between prenatal NO2 exposure and FL [190, 210, 211]. The majority of studies were conducted in European countries (Spain, Netherlands, and Scotland) followed by Asian countries (China and South Korea), with one study in North America (USA), and one study in Oceania (Australia). Study periods were conducted over 20 years (1993-2016) and mean NO2 levels ranged from 9.8 to 61.2 micrograms per meter cubed (μg/m 3 ). Wang et al. found NO2 from 0-22 weeks gestation estimated by land-use regression models at the home addresses was significantly associated with FL at the second trimester (𝛽 = -0.24mm per 10 μg/m 3 ; 95% CI -0.37, -0.12) [211]. Iñiguez et al. found elevated residential prenatal NO2 exposure (cut at the median of 38 μg/m 3 ) and estimated by temporally adjusted land-use regressions models was significantly associated with FL at 20 and 32 weeks, and with growth from 12-20 weeks gestation [206]. Van den Hooven et al., found prenatal traffic-related NO2 from conception until the second trimester ultrasound to be associated with FL at the second trimester (𝛽 = -0.02mm per 1 μg/m 3 increase; 95% CI -0.03, -0.01) and prenatal NO2 from conception until the third trimester ultrasound to be associated with FL at the third trimester (𝛽 =-0.02mm per 1 μg/m 3 increase; 95% CI -0.04, -0.01). Traffic-related NO2 concentrations at the residential locations were estimated using dispersion methods models [210]. Common covariates controlled for in models looking at this association included gestational age 47 at time of scan, season of conception, mother’s age, socioeconomic status (SES) measured as income and/or education, maternal race, pre-pregnancy BMI, prenatal smoking, supplement intake (i.e. folate) and fetal sex. Only one study reported a significant association from multipollutant adjusted models [208]. Prenatal NO2 exposure and AC has been assessed in eight studies, which overlap with the studies discussed above [190, 203-205, 207-209, 211]. Four of the studies found prenatal NO2 exposure to be significantly inversely related with AC; however, three of these come from the INMA Spanish project [203, 206, 211, 212]. Aguilera et al., Iñiguez et al. (2012) and Iñiguez et al. (2016) all conducted similar studies using subsets of the overall Spanish INMA project with outdoor NO2 estimated by temporally adjusted land-use regression models at the home address, with AC measured at 12-, 20-, and 34-weeks gestation and modeled as a mean percent change in standard deviation scores. All three studies found prenatal NO2 levels to have influences on decreased AC between 20-34 weeks [190, 203, 212]. Wang et al. found NO2 exposure from 0-22 weeks gestation to be significantly associated with AC at the second trimester (𝛽 = -1.02mm per 10 μg/m 3 ; 95% CI -1.57, -0.48) [211]. Common covariates were similar to what was listed prior, and only one study considered multipollutant modeling when looking at this association [208]. Prenatal outdoor NO2 exposure and BPD have been assessed in seven studies, which overlap with the studies discussed above [190, 203-205, 208, 209, 212], and five have found significant results [203, 206, 208, 209, 212]. Aguilera et al., Iñiguez et al. (2012) and Iñiguez et al. (2016), discussed above, found prenatal NO2 to be associated with decreased BPD growth between 20-34 weeks [190, 203, 212]. Lamichhane et al. estimated NO2 using land-use regression models based on home addresses and BPD measured in second and third trimester ultrasound scans in 648 South Korean women. NO2 was significantly associated with decreased BPD in the second 48 trimester (𝛽 = -0.26 mm per 10 μg/m 3 ; 95% CI -0.42, -0.11) and in the third trimester (𝛽 = -0.31 mm per 10 μg/m 3 ; 95% CI -0.59, -0.03). Second trimester results remained significant when additionally adjusting for PM10 [208]. Ritz et al. estimated outdoor residential NO2 concentrations using community based central monitoring measurements from the US EPA’s Air Quality System in the Children’s Health Study (1992-1996) in 566 women living in Los Angeles. BPD was measured at 18-20 weeks, 28-30 weeks, and 35-37 weeks. In the third period of measurement, NO2 was significantly inversely associated with BPD (𝛽 = -1.0mm per IQR, 95% CI -1.66, -0.34) [54]. Covariates for this association were similar to those previously listed for FL, and only one study considered multipollutant modeling approaches [208]. Six studies have assessed prenatal NO2 with EFW, which overlap with the studies discussed above [203, 206-208, 210, 211], and five have found significant inverse results, all within the third trimester of pregnancy [203, 206, 207, 210, 211]. Aguilera et al., Iñiguez et al. (2012) and Iñiguez et al. (2016), discussed above, found prenatal NO2 to be associated with decreased EFW between 20-34 weeks [190, 203, 212]. Van den Hooven, discussed above, found mothers in the highest quartile of prenatal NO2 exposure had fetuses that weighed -14.2 grams (95% CI -28.0, -0.5) less in the third trimester compared to those in the first quartile [210]. Wang et al., discussed above, found that for each 10 μg/m 3 increase in NO2 exposure, EFW in the third trimester decreased by 7.8 grams (95% CI -11.59, -4.08) [211]. Common covariates within this association were similar to those discussed above, and only one study utilized multipollutant models [208]. Four studies have assessed prenatal NO2 with fetal HC [205, 209-211], which overlap with previously discussed studies, and two have found significant results [210, 211]. Wang et al., discussed above, found prenatal NO2 to significantly decrease HC in both the second trimester (𝛽 = 49 -1.07mm per 10 μg/m 3 ; 95% CI -1.60, -0.54) and third trimester (𝛽 = -0.71mm per 10 μg/m 3 ; 95% CI -1.37, -0.06) [211]. Van den Hooven, discussed above, found a significant trend in NO2 exposure and HC at the third trimester where for each 1 μg/m 3 increase in NO2, HC decreased by 0.12 mm (95% CI -0.17, -0.06) [210]. Prenatal PM10 Exposure and Fetal Growth Six studies have assessed prenatal PM10 exposure and fetal FL [204, 205, 208-210, 213], and two have found significant inverse results [205, 210]. Most studies were done in European countries (Netherlands and Scotland), or Asian countries (China and South Korea) followed by one study in the US, and one in Australia with study periods beginning in 1993 up to 2012. A large range of exposure levels of PM10 was reported from 13-141 μg/m 3 . Van den Hooven et al. found PM10 measured from conception until the third trimester by dispersion methods at home addresses in the Netherlands to be significantly associated with third trimester FL when assessed in quartiles (𝛽 = -0.15mm Q4 vs Q1; 95% CI -0.3, 0.00) [210]. Hansen et al. estimated PM10 based on daily values from the closest monitoring site from the mother’s postcode in Australia for the first four months of pregnancy, which were averaged in 30-day periods. FL was measured in ultrasound scans that were taken between 13- and 26-weeks gestation. Results found that prenatal PM10 from 0-30 days gestation to be significantly associated with FL at mid-pregnancy modeled as a mean change (𝛽 =-0.28 per 5 μg/m 3 ; 95% -0.48, -0.08) and PM 10 91-120 days with mid pregnancy FL (𝛽 = -0.23mm per 5 μg/m 3 ; 95% CI -0.42, -0.04). Results remained consistent and statistically significant when additionally adjusted for O3 [205]. Common covariates included in models looking at PM10 and FL included gestational age at time of scan, mother’s age, pre-pregnancy BMI, season of conception, parity, SES (education and/or household income), sex of fetus, 50 supplement intake, and prenatal smoking. Two studies utilized models that adjusted for multiple pollutants [205, 208]. Five studies, which overlap with those discussed above, have assessed prenatal PM 10 exposure and AC [204, 205, 208, 209, 213], with two studies finding significant inverse results [205, 208]. Lamichhane et al. estimated residential PM10 using land-use regression models, and AC was measured in second and third trimester ultrasound scans in 648 South Korean women. Within adjusted models, a non-significant inverse relationship was observed with PM10 and AC, however, when authors additionally adjusted for prenatal NO2 exposure PM10 was significantly associated with AC in the third trimester (𝛽 = -3.0mm per 10 μg/m 3 ; 95% CI -5.7, -0.38) [208]. Hansen et al., discussed above, found prenatal PM10 between 91-120 days gestation to be significantly associated with mid-pregnancy AC (13-26 weeks gestation) 𝛽 = -0.78mm per 5 μg/m 3 (95% CI -1.49, -0.08), and results remained significant when additionally adjusting for O3 [205]. Covariates commonly used in multivariable models were the same seen in PM 10 and FL associations. Three of the five studies included multipollutant models [205, 208, 213]. Five studies, which overlap with those previously discussed, have assessed prenatal PM10 exposure and BPD [204, 205, 208, 209, 213], and three have found significant inverse relationships [204, 208, 213]. Lamichhane et al., discussed above, found PM10 to have a significant influence on decreased BPD at the third trimester (𝛽 = -0.31mm per 10 μg/m 3 ; 95% CI -0.59, -0.03), however, after additionally adjusting for prenatal NO2, results were insignificant [208]. Clemens et al. estimated residential, annual PM10 concentrations with spatial resolution of 1km x 1km at the centroid of the mothers’ postcode of residence. Within 13,775 pregnant Scottish women, authors found PM10 to be associated with 3 rd trimester BPD (𝛽 = -0.16 difference in SD score per 10 μg/m 3 ; 95% CI -0.30, -0.02), but not second trimester BPD [204]. Zhao et al. estimated PM10 exposures 51 in Lanzhou, China, incorporating the mother’s home and work addresses for day of conception until day of ultrasound, one month prior to ultrasound, two weeks prior to ultrasound, and one week prior to ultrasound. The study population had high levels of exposure (N=8877, mean PM10=140.8 μg/m 3 ). BPD was modeled as standardized z-scores, and PM10 from conception day until ultrasound scan was found to be significantly inversely associated with BPD after adjustment for covariates including SO2 and NO2 (𝛽 = -0.02 per 10 μg/m 3 ; 95% CI -0.03, -0.01) [213]. Models that assessed prenatal PM10 and BPD included similar covariates discussed for FL and two studies included models with multiple pollutants [208, 213]. Three studies have assessed prenatal PM10 on EFW, and all have found significant results, although in differing directions [208, 210, 214]. Van den Hoover, discussed above, found PM10 exposure to positively influence EFW at the second trimester (𝛽 =3.8 grams Q4 vs Q1; 95% CI 0.8, 6.8) after adjusting for gestational age and noise exposure at ultrasound, BMI, height, parity, ethnicity, education, folic acid, supplementation use, prenatal smoking, alcohol use, paternal height, and fetal sex. The direction of PM10 on EFW switched directions in the third trimester, however, this association was not statistically significant (𝛽 = -11.0 grams Q4 vs Q1; 95% CI - 23.2, 1.2). Models did not assess the influence of co-exposure to multiple pollutants [210]. Lamichhane et al., discussed above, found PM10 to inversely influence EFW at the third trimester in single pollutant adjusted models, but results were not statistically significant. Covariates included gestational age at scan, pre-pregnancy weight, height, parity, education, urine cotinine levels, dietary folate, fetus sex, and study area. After additionally adjusting for NO2, PM10 significantly influenced EFW at the third trimester (𝛽 = -62.3 grams per 10 μg/m 3 ; 95% CI -118.0, -6.5) [208]. Lin et al. conducted a study in Beijing, China, from a hospital cohort of 18,863 women. PM10 was estimated from daily concentrations at each residential address using inverse distance 52 weighting approaches and an overall average was computed from date of last menstrual period to date of delivery for each participant. Mean PM10 concentrations were higher compared to other cohorts (116.49 μg/m 3 ). EFW was estimated using the Hadlock et al. formula and was measured longitudinally up to three scans per participant and was modeled as a gestational-age and gender- adjusted Z-score. After adjusting for maternal age, ethnicity, employment status, education, parity, year of conception, ambient temperature and normalized difference vegetation index, for each 10μg/m 3 increase in PM10 exposure, EFW decreased by -0.030 (95% CI -0.043, -0.017) [214]. Three studies have assessed prenatal PM10 and HC [205, 209, 210], with two finding significant inverse results [205, 210]. Hansen et al., discussed above, found prenatal PM10 91-120 days gestation to be significantly associated with mid pregnancy HC (𝛽 = -1.02mm per IQR; 95% CI -1.78, -0.26) after adjusting for fetal sex, gestational age at scan, mother’s SES, temperature exposure, seasonality and long-term trend. When models were additionally adjusted for O3, results remained significant (𝛽 = -1.05mm per IQR; 95% CI -1.83, -0.28) [205]. Van den Hoover, discussed above, found PM10 exposure to negatively influence HC at the third trimester (𝛽 = - 0.18mm per 1 μg/m 3 ; 95% CI -0.24, -0.12) after adjusting for gestational age and noise exposure at ultrasound, BMI, height, parity, ethnicity, education, folic acid, supplementation use, prenatal smoking, alcohol use, paternal height, and fetal sex. No models within this study assessed the influence of multiple pollutants concurrently [210]. Prenatal PM2.5 Exposure and Fetal Growth Three studies have assessed prenatal PM2.5 exposure on FL [204, 215, 216], while only one has found PM2.5 to significantly influence lower FL [215]. These studies were conducted in Scotland, China, and the US with mean PM2.5 exposures ranging from 7.2-81.5 μg/m 3 , which were 53 measured from 1993-2017. Cao et al. measured PM2.5 from weekly average values from 53 monitors and utilized the inverse distance weighted method based on residential addresses to model exposure levels which were modeled as the average exposure from first gestational week to week before ultrasound (gestational age at scan: 14-40 weeks) within 7,965 Chinese women. Generalized estimating equations were used to account for multiple measurements and PM2.5 was significantly associated with FL (𝛽 = -5.82mm per 10μg/m 3 ; 95% CI -6.71, -4.92) after adjusting for gestational age at scan, fetus sex, pregnancy hypertension, gestational diabetes, mother’s age, parity, gravity, medical insurance and season of conception. Results remained significant in multipollutant models additionally adjusting for SO2, NO2, PM10, and O3 [215]. Three studies to date have assessed prenatal exposure to PM2.5 and AC [204, 215, 216] and two have found significant inverse results [215, 216]. Cao et al, discussed above, found that for each 10μg/m 3 in PM2.5 AC decreased by -5.82 mm (95% CI -6.71, -4.92), and results remained significant when additionally adjusting for SO2, NO2, PM10, and O3 (𝛽 = -5.57mm per 10μg/m 3 ; 95% CI -9.06, -2.08) [215]. Leung et al. is the only study conducted within the United States to assess prenatal PM2.5 exposure and fetal growth [216]. The study population was predominantly non-Hispanic white women, although there were mothers of color included (17% Black, 9% Hispanic, 10% Asian) and participants came from a private hospital in Eastern Massachusetts. Abdominal circumference was measured across varying weeks of gestation but was analyzed as scans prior to 24 weeks (considered “anatomic scans”) or scans above 24 weeks (considered “growth scans”) and were modeled as z-scores. PM2.5 was measured using a spatiotemporal model and participant addresses and were averages from conception until week 16 of gestation as well as an average from date of conception until date of assessment. The mean PM2.5 exposure level across pregnancy was 7.32 μg/m 3 . The authors found that averaged PM2.5 from conception until 16 weeks 54 gestation was significantly inversely associated with AC at <24 weeks (𝛽 = -0.15 per 5 μg/m 3 ; 95% CI -0.28, -0.01) and at >24 weeks gestation (𝛽 = -0.16 per 5 μg/m 3 ; 95% CI -0.31, -0.02). Only AC taken at taken scans <24 weeks gestation remained significant in overall average models (𝛽 = -0.12 per 5 μg/m 3 ; 95% CI -0.24, -0.00). Models were adjusted for maternal age, race, education, insurance type, parity, fetal sex, date of conception, temperature and area deprivation index. Multipollutant models were not included [216] . Three studies have accessed prenatal PM2.5 exposure and BPD, and all have found significant inverse results [204, 215, 216]. Cao et al., discussed above, found that for each 10μg/m 3 in PM2.5, BPD decreased by -5.82 mm (95% CI -6.88, -4.75), and results remained significant when additionally adjusting for SO2, NO2, PM10, and O3 (𝛽 = -5.77mm per 10μg/m 3 ; 95% CI -6.84, -4.69) [215]. Leung et al., discussed above, found that averaged PM2.5 from conception until 16 weeks gestation was significantly inversely associated with BPD at <24 weeks (𝛽 = -0.19 per 5 μg/m 3 ; 95% CI -0.31, -0.06) and at >24 weeks (𝛽 = -0.15 per 5 μg/m 3 ; 95% CI -0.26, -0.04). Only BPD measured at scans <24 weeks gestation remained significant in overall average models (𝛽 = -0.15 per 5 μg/m 3 ; 95% CI -0.26, -0.04) [216]. Clemens et al. estimated residential, annual PM2.5 concentrations with spatial resolution of 1km x 1km at the centroid of the mothers’ postcode of residence. Within 13,775 pregnant Scottish women, authors found PM2.5 to be associated with 3 rd trimester BPD (𝛽 = -0.43 difference in SD score per 10 μg/m 3 ; 95% CI -0.63, -0.24), but not second trimester BPD after adjusting for parity, age, sex, smoking, social class, maternal height and weight [204]. Two studies to date have assessed prenatal PM2.5 exposure on EFW [214, 215]. Cao et al., discussed above, found that for each 10μg/m 3 increase in PM2.5, EFW decreased by -14.45 grams (95% CI -16.96, -15.72), and results remained significant when additionally adjusting for SO 2, 55 NO2, PM10, and O3 (𝛽 = -14.53grams per 10μg/m 3 ; 95% CI -15.81, -13.34) [215]. Lin et al. conducted a study in Beijing, China from a hospital cohort of 18,863 women. PM2.5 was estimated from daily concentrations at each residential address using inverse distance weighting approaches and an overall average was computed from date of last menstrual period to date of delivery for each participant. Mean PM2.5 concentrations were higher compared to other cohorts (81.52 μg/m 3 ). EFW was estimated using the Hadlock et al. formula and was measured longitudinally up to three scans per participant and was modeled as a gestational-age and gender-adjusted z-score. After adjusting for maternal age, ethnicity, employment status, education, parity, year of conception, ambient temperature and normalized difference vegetation index, for each 10μg/m 3 increase in PM2.5 exposure, EFW decreased by -0.031 (95% CI -0.047, -0.016) [214]. Only one study, Leung et al., has accessed the association between prenatal PM2.5 exposure and HC measured by ultrasound, and no significant associations were observed [216]. Prenatal O3 Exposure and Fetal Growth Two studies have accessed prenatal O3 and FL, with neither finding significant relationships [205, 209]. Hansen et al. estimated O3 based on daily values from the closest monitoring site from the mother’s postcode in Australia for the first four months of pregnancy, which were averaged in 30-day periods. The study period was 1993-2003 and FL was measured in ultrasound scans that were taken between 13- and 26-weeks gestation. Covariates included fetal sex, gestational age, mother’s age, mother’s SES, concurrent temperature exposures, seasonality and long-term trend. No significant findings were seen with O3 and FL in this study, although an inverse relationship was suggested in all time periods [205]. Ritz et al. estimated outdoor residential O3 concentrations using community based central monitoring measurements from the 56 US EPA’s Air Quality System in the Children’s Health Study (1992-1996) in 566 women living in Los Angeles and FL was measured at 18-20 weeks, 28-30 weeks, and 35-37 weeks. No significant association was seen with O3 and FL after adjusting for gestational period of pregnancy for ultrasound, maternal age, race/ethnicity, education level, parity, prenatal care payment, marital status, maternal smoking before each ultrasound visit (yes/no), maternal infections before each ultrasound visit, maternal height, maternal pre-pregnancy weight, weight gain in each pregnancy period, presentation (breech vs. vertex/transverse), and sonographer [209]. Neither study included results for multipollutant models. Only two studies have assessed prenatal O3 and AC, which are the same studies discussed above[205, 209], although only one found significant inverse associations [205]. Hansen et al. found that for O3 exposure 1-30 days of gestation to be significantly associated with decreased AC (𝛽 =-1.42mm per 5 μg/m 3 ; 95% CI 2.74, -0.09). Results remained significant when additionally adjusting for PM10 [205]. The same studies discussed above are the only ones to date to have accessed prenatal O3 with BPD, and neither found significant results [205, 209]. No studies to date have assessed prenatal O3 exposure and EFW. The same studies discussed above are the only ones to date to have accessed prenatal O3 with HC, and neither found significant results [205, 209]. Gaps Within the Literature Overall, existing research suggests that prenatal exposure to air pollutants negatively affects several measures of fetal growth. However, this review has served as a method to identify critical gaps within the current literature. Notably, no studies to date have utilized distributed lag 57 model (DLM) techniques to assess finer windows of exposure on in utero fetal growth. Within the current literature, the most standard measure for exposure assessment is an overall average either for the duration of pregnancy or from date of conception to date of ultrasound. DLMs simultaneously use finer windows (i.e. weeks) to identify susceptible periods of exposure that may be critical for identifying subtle, but important, influences on fetal growth. Although some studies did calculate different mean exposure periods, such as by trimesters, months, or additional averages (i.e. 16 weeks) in separate models, this does not capture the true variability of how exposure levels change across pregnancy or address the correlation of exposure levels across time. Current work also only explores the influence of prenatal exposure levels and using the DLM framework would allow both preconception and gestational exposure levels to be assessed for impacts on in utero fetal growth. Additionally, most studies have been conducted in homogeneous populations predominately in European and Asian countries. Very little work has been conducted in North America, and only two studies have included participants from ethnic and racial minority populations, Ritz et al. (31% Hispanic, 42% Black) [209] and Leung et al. (17% Black, 9% Hispanic, 10% Asian) [216]. Given that it is well known that race and ethnicity are significantly associated with disproportionate exposure to outdoor air pollution within the US [186, 187, 217], as well as women of color having higher rates of adverse birth outcomes compared to non-Hispanic white women [218], it is important to better understand the influence of outdoor air pollution on fetal growth in utero within health disparities populations. Future studies assessing outdoor air pollution on fetal growth should also incorporate O3 and PM2.5, as the literature on these pollutants and fetal biometry measurements are underrepresented compared to NO2 and PM10. Additionally, some studies did not have access to 58 residential addresses and completed air pollutant exposure assessment based on closest monitoring sites and postal codes [204, 205]. Future studies should incorporate more precise exposure assessment techniques, especially those that capture movement of residences across gestation. Another large gap in the literature is that few studies incorporated multipollutant models, and single pollutant exposure models alone do not capture the complex mixture of pollutants that humans are exposed to which can interdependently influence fetal growth. 59 CHAPTER THREE Overview of the MADRES Study Cohort Description All three of the studies within this dissertation were conducted within the Maternal And Developmental Risks from Environmental and Social Stressors (MADRES) pregnancy cohort. MADRES is an ongoing prospective study that began recruitment in November 2015 and will consist of 1,000 predominately low-income, Hispanic women residing in urban Los Angeles at the end of recruitment. Bastain et al. previously described the study design and protocol in detail [219]. Pregnant mothers are recruited from four prenatal clinic sites that typically serve medically underserved populations. These include two community health clinics, one county hospital prenatal clinic, and one private obstetrics and gynecology practice. In addition, a small segment of the cohort is from community recruits, which are women with similar demographics to the patients seen within the clinic sites. Eligibility for participants at the time of recruitment includes being less than 30 weeks pregnant, at least 18 years of age, and a fluent speaker of English or Spanish. Exclusion criteria for study participation includes multiple gestation, having a physical, mental, or cognitive disability that would prevent participation or ability to provide consent, current incarceration, and HIV positive status. To date (May 2022), 733 recruited mothers have delivered their infants within the MADRES cohort. Their demographics and variables related to study design are shown in Table 1 (mothers) and Table 2 (infants). 60 Table 1. Maternal Demographics (N=733) Characteristic Mean (SD) / N (%) Study Design Recruitment Site Eisner LAC+USC USC OB-GYN South Central Clinic Community Recruit 514 (70.1%) 132 (18.0%) 58 (7.9%) 24 (3.3%) 5 (0.7%) Study Entry <20 weeks gestation 20-29 weeks gestation 545 (74.3%) 188 (25.7%) Maternal Age at recruitment, years 28.6 (6.7) Race/Ethnicity (N=690) Hispanic Non-Hispanic Black Non-Hispanic White Non-Hispanic Other Non-Hispanic, multiracial 542 (78.6%) 85 (12.3%) 38 (5.5%) 16 (2.3%) 9 (1.3%) Household Income (N=691) <$15,000 $15,000-$29,999 $30,000-$49,999 $50,000-$99,999 >$100,000 Reported Don’t Know 142 (20.5%) 169 (24.5%) 76 (11.0%) 38 (5.5%) 35 (5.1%) 231 (33.4% Education (N=690) <12 th Grade Completed High School Some college/technical school Completed 4 years of college Some graduate training post college 177 (25.6%) 217 (31.5%) 184 (26.7%) 73 (10.6%) 39 (5.6%) Pre-pregnancy Body Mass Index (BMI) (N=722) Underweight (<18.5 BMI) Normal Weight (18.5-25 BMI) Overweight (25-29 BMI) Class 1 Obese (30-34 BMI) Class 2 Obese (35-39 BMI) Class 3 Obese (40+ BMI) 18 (2.5%) 220 (30.5%) 222 (30.7%) 151 (20.9%) 65 (9.0%) 46 (6.4%) Marital Status (N=598) Living together Married Never married, single Divorced or separated Declined to answer 237 (39.6%) 188 (31.4%) 141 (23.6%) 19 (3.2%) 13 (2.2%) Country of Origin (N=608) USA Mexico Other 332 (54.6%) 158 (26.0%) 118 (19.4%) 61 Table 2. Infant Characteristics (N=733) Characteristic Mean (SD) / N (%) Birth weight, grams (N=724) 3263.1 (522.2) Born Low Birth Weight (<2500 grams) (N=705) 42 (25.8%) Gestational Age at Birth, weeks 38.9 (1.8) Born premature (<37 weeks) 73 (9.9%) Sex (N=730) Female 366 (50.1%) Birth Order (N=607) First Born Second Born Third Born Fourth Born Fifth Born Sixth + Born 224 (36.9%) 193 (31.8%) 109 (18.0%) 50 (8.2%) 19 (3.1%) 12 (2.0%) Relevant Methodology Data are collected via in person visits, telephone questionnaires, and medical record abstractions. In person visits consist of interviewer-administer questionnaires, anthropogenic measurements, and biospecimen collections. Points of contact either in-person or by telephone span throughout pregnancy and until the child’s fifth year of life and include the first, second and third trimester of pregnancy, 7-14 days after birth, 1, 3, 8, 12, 18-month(s) post birth and 2, 2.5, 3, 3.5, 4, 4.5 years post birth. This relevant methodology section will focus on the two prenatal environmental exposures (PFAS and ambient air pollution), birth outcomes (infant birthweight and gestational age at birth), in utero fetal growth measurements, and covariates used in the three dissertation studies. Prenatal PFAS Blood Serum Concentrations The environmental prenatal exposure of PFAS concentrations were measured in 359 MADRES participants from a single maternal blood sample collected at either the initial study visit or at the in-person third trimester study visit (≥ 28 weeks gestation). Only one blood sample per mother was measured for PFAS concentrations and gestational age at time of sample ranged 62 from 5.7 weeks to 38.3 weeks (mean = 20.8 ±9.3 weeks). Samples were collected between December 2015 to February 2019 using red top 10-milliliter serum tubes which were then processed and aliquoted into appropriated labeled 0.5 ml cryovials before being stored at -80° Celsius. Storage occurred prior to shipment to the analytical laboratory, which was the Wadsworth Center’s Human Health Exposure Assessment Resource (WC-HHEAR) laboratory at NYU Langone Medical Center (Dr. Kannan’s Laboratory). Samples preparation, extraction, instrumental analysis using ultrahigh-performance liquid chromatography tandem mass spectrometry, and data quality control/review were conducted in Spring 2019. A total of 14 PFAS analytes were measured in the blood serum. Prenatal Ambient Air Pollution Exposure The environmental exposure of prenatal ambient air pollution was assigned based on daily residential histories that were assembled and geocoded for each participant using residential address and occupancy dates collected with residential history forms and prospective address confirmation data was collected at every contact point. These capture all residential mobility or moves and form the basis of all geospatial exposure assignments. Daily ambient air pollutant concentrations of PM2.5, PM10, O3 (8-hour max) and NO2, were then assigned using inverse- distance-squared weighted spatial interpolation from the United States Environmental Protection Agency (EPA) Air Quality System monitors. For PM2.5, PM10, and NO2, 24-hour daily averages were used, while the 8-hour daily maximum was used for O3. Daily averages capturing air pollutant exposure for this current work ranged from 12 weeks prior to conception until the third trimester of pregnancy. 63 Infant Birthweight and Gestational Age at Birth For the full MADRES sample, infant birthweight was abstracted from the electronic or paper medical record of the delivery as a primary source (99.9%) and if unavailable self-reported by the mother 7-14 days post birth via a phone interviewer-administered questionnaire as a secondary source (n=1). Gestational age at birth was calculated and standardized using a hierarchy of clinically optimal methods [220]. The preferred method was by using a measurement of crown- rump length from a first trimester (<14 weeks gestation) ultrasound (62%), followed by a measurement of biparietal diameter from a second trimester (14 week to <28 weeks gestation) ultrasound (26%). If early ultrasound data were missing, gestational age at birth was estimated from a physician’s best clinical estimate from obstetrical medical records (11%), or only estimated from last menstrual period (LMP), which was deemed least desirable (<1%). Birthweight for gestational age z-scores that are infant sex or parity specific were calculated using the formula presented by Aris et al. [221]. In Utero Fetal Growth Parameters Ultrasound measured fetal growth parameters came from two sources. The first was scans that were abstracted from all available prenatal ultrasound visits. MADRES currently has information on a total of 2,436 ultrasound scans that range from 3-42 weeks of gestation. The second source is a study-measured ultrasound by a licensed sonographer during the in-person third trimester visit (≥ 28 weeks gestation), which added an additional 285 scans (mean gestational age at scan=32 ±2 weeks). Ultrasound scans were conducted either at Keck Hospital by two technicians (August 2016–August 2018, n=68 scans) or in the MADRES study clinic by one technician (August 2018–March 2020, n=217 scans). The ultrasound scan information included within the current dissertation studies are the following four growth parameters: abdominal 64 circumference, head circumference, femur length, and biparietal diameter, which were measured in a standardized fashion [222]. Estimated fetal weight was calculated using the above parameters and the Hadlock et al. formula [223]. Covariates Covariates including maternal race/ethnicity, age at enrollment, highest level of education, annual household income, parity (birth order of the child), personal smoking during pregnancy, dietary fish consumption during pregnancy, pre-pregnancy weight and country of origin were collected during the prenatal life stage through interviewer-administered questionnaires. Medical health conditions such as diabetes (chronic or gestational) or hypertension status (chronic or gestational) have been abstracted from maternal electronic medical records (EMR). Sex of the fetus/infant is primarily abstracted from the EMR (97.5%), followed by the birth information form (1.5%), or proxy reported by the mother (1%). Pre-pregnancy BMI was calculated using self- reported pre-pregnancy weight and study-measured height at the first visit via a stadiometer (Perspectives enterprises model PE-AIM-101). MADRES Funding Sources MADRES Center (grant #s P50ES026086, 83615801, P50MD015705) funded by the National Institute of Environmental Health Sciences (grant # R01ES027409), the National Institute for Minority Health and Health Disparities and the Environmental Protection Agency; the Southern California Environmental Health Sciences Center (grant # 5P30ES007048) funded by the National Institute of Environmental Health Sciences, and the Life course Approach to Developmental Repercussions of Environmental Agents on Metabolic and Respiratory health (LA DREAMERs) (grant #s UH3OD023287) funded by the National Institutes of Health Office of the Director ECHO Program. 65 CHAPTER FOUR Study 1: Prenatal Perfluorooctanoic Acid (PFOA) Exposure is Associated with Lower Infant Birthweight Within the MADRES Pregnancy Cohort Abstract Introduction Per- and polyfluoroalkyl substances (PFAS) are persistent synthetic chemicals found in many household products that can cross the placenta during pregnancy. We investigated whether exposure to PFAS during pregnancy were associated with infant birth outcomes in a predominately urban Hispanic population. Methods Serum concentrations of perfluorooctanoic acid (PFOA), perfluorooctanesulfonic acid (PFOS), perfluorohexanesulfonic acid (PFHxS), perfluorononanoic acid (PFNA), and perfluorodecanoic acid (PFDA) were measured in 342 prenatal biospecimens (mean gestational age: 21±9 weeks) from participants in the ongoing Maternal And Developmental Risks from Environmental and Social Stressors (MADRES) cohort. PFAS compounds were modeled continuously or categorically depending on percentage of detected samples. Birth outcomes assessed were birthweight, gestational age at birth, and birthweight for gestational age (BW-for-GA) z-scores that accounted for parity or infant sex. Single pollutant and multipollutant linear regression models were performed to evaluate associations between PFAS exposures and birth outcomes adjusting for sociodemographic, perinatal, and study design covariates. Results Maternal participants (n=342) were on average 29±6 years old at study entry and were predominantly Hispanic (76%). Infants were born at a mean of 39±2 weeks gestation and weighed 66 on average 3,280±523 grams. PFOS and PFHxS were detected in 100% of samples while PFNA, PFOA, and PFDA were detected in 70%, 65%, and 57% of samples, respectively. PFAS levels, with the exception of PFHxS, were lower in this cohort than in comparable cohorts. Women with detected levels of PFOA during pregnancy had infants that weighed on average 119.7 grams less (95% CI -216.7, -22.7) than women with undetected levels of PFOA when assessed in adjusted single pollutant models. PFOA results were also statistically significant in BW-for-GA z-score models that were sex or parity specific. Within models that mutually adjusted for five detected PFAS compounds, PFOA results remained comparable; however, the association was only significant in BW-for-GA z-scores that were parity specific (𝛽 = -0.3; 95% CI -0.6, -0.01). We did not find significant adjusted associations with the remaining PFAS concentrations and the birth outcomes assessed. Conclusion Prenatal exposure to PFOA was associated with lower birthweight in infants, suggesting that exposure to these chemicals during critical periods in development may have important implications for children’s health. Introduction Per and polyfluoroalkyl substances (PFAS) are a class of ubiquitous synthetic chemicals that have bioaccumulating properties with long half-lives and are commonly found in household products due to their water and stain resistant qualities. Across the world, these persistent chemicals have been found in humans, marine and land animals, and soil and water in the environment [224]. Within this class of chemicals, most research has focused on perfluorooctanoic acid (PFOA) and perfluorooctanesulfonic acid (PFOS); however, there are over 4,700 compounds that are considered PFAS [225]. Animal and in vitro studies have found PFOA and PFOS to have 67 toxic effects on the liver, kidneys, reproductive system, cardiovascular system and endocrine system [226-229]. Prenatal exposure to PFOA and PFOS have also consistently led to lower birthweight and offspring mortality in mice and rats [230, 231]. Although PFOA, PFOS, and PFAS with similar chemical structures have been phased out of manufacturing in the United States (US) [232], products containing them can still be imported. In addition, due to their long half-lives (PFOA 4 years and PFOS 5 years [233]) the substances are still commonly found in the blood serum of the US population, including in children born after the phase out in 2002 [234, 235]. The most common exposure routes of PFAS to humans include ingestion through drinking water or contaminated food (i.e. fish or food stored in packaging containing PFAS transferred into food), inhalation of indoor air and dust, and occupational exposure through inhalation or dermal contact [236-241]. Exposure to these substances may be particularly critical during fetal development due to their ability to cross the placenta during pregnancy [242]. The placenta has been shown to be vulnerable to environmental insults and is a common target of PFAS where it has been shown to accumulate across gestation via maternal circulation [242-244]. A consensus on the exact biological mechanism for how PFAS influences fetal growth has not been met; however, hypotheses including impacts on bone development, maternal hormone disruption - particularly sex and thyroid hormones - and adverse effects on placental development and function have been proposed [245-252]. Several recent meta-analyses and systematic literature reviews suggest that prenatal PFOA and PFOS exposures are associated with lower infant birthweight [231, 253-256]. However, many individual studies have failed to detect significant associations. Additionally, several biases have been discussed, especially for studies that measure concentrations late in 68 pregnancy or at birth, due to biological and physiological changes that occur in later pregnancy, such as increased glomerular filtration [255]. Epidemiological studies have also found statistically significant associations between other PFAS compounds including perfluorohexanesulfonic acid (PFHxS), perfluorononanoic acid (PFNA), and perfluorodecanoic acid (PFDA) with infant birthweight [257-262]; however, these relationships have been assessed in far fewer studies than with PFOA and PFOS. Many recently published studies analyzing these associations used specimens collected in the early 2000s, which likely do not represent current exposure levels due to the phase outs and use of replacement compounds that have occurred in recent decades [259, 263-265]. Better understanding of the association between prenatal PFAS exposure, particularly compounds other than PFOA and PFOS, and infant birthweight is important since low birthweight, defined as <2,500 grams, is a strong predictor of childhood obesity and subsequent metabolic health morbidities, such as type 2 diabetes later in life [266-268]. Within the US, pregnant persons identifying as Black or Hispanic have higher rates of infants born with low birthweight compared to their non-Hispanic white counterparts [269]. PFAS concentration levels also differ by race and ethnicity. The National Health and Nutrition Examination Survey (NHANES) found non-Hispanic whites to have higher concentration levels in reference to Blacks or Hispanics [234]. However, few epidemiologic studies have investigated the influence of PFAS exposure on birth outcomes among racially and ethnically diverse populations, specifically within Hispanic populations. The objective of this study was to determine if maternal serum PFAS concentrations measured in pregnancy were associated with infant birthweight and birthweight for gestational age (BW-for-GA) z-scores in a prospective pregnancy cohort representing a structurally marginalized 69 population. Additionally, we examined the relationship between prenatal PFAS serum concentrations and gestational age at birth. Methods Study Sample Participants included in this study were drawn from the ongoing Maternal And Developmental Risks from Environmental and Social Stressors (MADRES) pregnancy cohort, which is predominately comprised of Hispanic women and children who reside in urban Los Angeles, CA, USA. Methods of the MADRES study and protocol have been described previously [270]. Briefly, participants were enrolled during pregnancy from four prenatal clinic sites. These included two community health clinics, one county hospital prenatal clinic, and one private obstetrics and gynecology practice. Eligibility for participants at the time of recruitment included: 1) less than 30 weeks pregnant; 2) at least 18 years of age; and 3) a fluent speaker of English or Spanish. Exclusion criteria for the study included: 1) multiple gestation; 2) having a physical, mental, or cognitive disability that would prevent participation or ability to provide consent; 3) current incarceration; or 4) HIV positive status. At study entry, informed consent and HIPAA authorization was obtained for medical record abstraction from all participants. The University of Southern California’s Institutional Review Board approved all study procedures. Blood serum samples were collected from 359 MADRES participants between late 2015 and February 2019 and PFAS concentrations were measured in Spring of 2019. Of the 359 participants with PFAS measurements, 347 participants had infant birthweight data available. We removed five participants who withdrew from the study prior to birth (two study dropouts, three miscarriages) as well as seven participants whose infants were missing birthweight. An additional five participants were removed due to missing race/ethnicity. A total of 342 mother-infant dyads 70 with PFAS concentrations, birth outcomes data, and key covariate information were included in the current study. The consort diagram depicting available data is shown in Figure 1. Figure 1. Consort Diagram of Included Mother-Infant Dyads Exposure Assessment of Prenatal PFAS Concentrations Blood samples were collected during pregnancy using red top 10 milliliter serum tubes at an in-person study visit (median gestational age: 19.1 weeks; range 5.7-38.3 weeks). Samples were processed and serum was aliquoted into appropriately labeled 0.5 ml cryovials before being stored in a -80°C freezer prior to shipment to the analytical laboratory. Samples were sent to the Wadsworth Center’s Human Health Exposure Assessment Resource (WC-HHEAR) laboratory at NYU Langone Medical Center (Dr. Kannan’s laboratory) for analysis. Fourteen PFAS namely, PFHxS, PFOS, PFOA, PFNA, PFDA, perfluorobutanesulfonic acid (PFBS), perfluoroheptanoic acid (PFHPA), perfluroundecanoic acid (PFUNDA), perfluorododecanoic (PFDODA), perfluorooctanesulfonamide (PFOSA), n-ethyl perfluorooctane sulfonamido acetic acid (NETFOSAA), n-methyl perfluorooctane sulfonamido acetic acid (NMFOSAA), perfluoro-n-pentanoic acid (PFPEA), and perfluorohexanoic acid (PFHxA) were analyzed. The method for the analysis of 14 PFAS in serum has been described 359 Participants with PFAS Measurements (n=359) Available Infant Birthweight (n=347) Infant birthweight not recorded (n=12) Final Sample: n=342 Missing data on key covariates (n=5) 71 elsewhere [271]. In brief, serum samples (0.25 mL) were aliquoted into 15 mL polypropylene (PP) tubes and spiked with 5 ng of 13 C-labeled internal standard (IS) mixture and 0.7 mL of 1% ammonium formate (w/v) in methanol (MeOH). The mixture was centrifuged for 5 min at 5000 rpm and the supernatant was collected and passed through Hybrid-SPE cartridge (Phospholipid, 30mg, 1cc, Sigma-Aldrich, St. Louis, MO, USA). The cartridges were conditioned with 1 mL of MeOH containing 1% ammonium formate (w/v). The samples were eluted through the cartridge and collected in a PP tube for LC-MS/MS analysis. The target analytes were quantified by isotopic dilution method and a 12-point calibration (at concentrations ranging from 0.02 to 100 nanograms per milliliter (ng/mL)) with the regression coefficient of ≥ 0.999 was used. A pure solvent (MeOH) and a mid-point calibration standard (5 ng/mL) were injected after every 10 samples to check for carry over of target chemicals and instrumental drift in sensitivity. Several procedural blanks were analyzed to monitor for contamination that can arise from reagents and materials used in sample preparation steps. For each batch of 100 samples, five duplicates of procedural blanks and QC spiked samples (water spiked with native standards at 5 ng for all analytes and IS) were processed. Duplicates of Standard Reference Material (SRM1958, NIST, Gaithersburg, MD, USA; IS spiked) containing certified values for PFHxS, PFOS, PFOA, and PFNA were analyzed. Trace levels of NETFOSAA (0.001- 0.047 ng/mL) and NMFOSAA (0.002-0.012 ng/mL) were found in procedural blanks, and the concentrations of these chemicals in samples were subtracted from blank values. Spiked sample and SRM1958 recoveries were in the ranges of 81.5-112% (RSD: ±4.3-8.7%) and 86.7-111% (RSD: ±2.1-6.8%), respectively. The limit of detection (LOD) of target analytes ranged from 0.02 to 0.05 ng/mL. 72 Five of the 14 PFAS were detected in at least 50% of MADRES participant samples and were included in succeeding data analysis. These included PFOS (100% detected), PFHxS (100% detected), PFNA (70% detected), PFOA (65% detected) and PFDA (57% detected). Birth Outcomes Infant birthweight was directly abstracted from electronic medical records (EMR) for all but one participant (99.7%). The remaining participant did not have a physician measurement of birthweight recorded on the EMR and birthweight was obtained from the participant via an interviewer-administered questionnaire with a MADRES staff member 7–14 days post-birth. Gestational age at birth was calculated and standardized using a hierarchy of methods [272]. A first trimester (<14 weeks gestation) ultrasound measurement of crown-rump length was deemed ideal and was used if available (59.7%). If unavailable, a second trimester (<28 weeks gestation) ultrasound measurement of fetal biparietal diameter was used (26.9%). If measurements from an early ultrasound were unavailable, gestational age at birth was calculated based on a physician’s best clinical estimate from EMR (13.2%). If none of the above were available, gestational age was calculated from the estimated last menstrual period (<1%). We evaluated birthweight modeled continuously and by calculating sex specific or parity specific BW-for-GA z-scores based on Aris et al. [273]. These z-scores reflect the current sociodemographic composition of the US using 2017 US natality files from obstetric estimated gestational age on singleton births, which replaced previous BW-for-GA z-scores estimated using self-reported last menstrual period [273]. Covariates Potential covariates were identified a priori from the literature and included participant demographics, aspects related to study design, and pregnancy and birth related variables. Study design variables were study recruitment site and gestational age at time of blood serum collection. 73 Participant demographic covariates included race/ethnicity, country of birth, highest attained education, annual household income, participant’s age at time of study enrollment, any reported personal smoking during pregnancy, and pre-pregnancy Body Mass Index (BMI). These variables were self-reported via interviewer-administered questionnaires in English or Spanish. Pre-pregnancy BMI was calculated using self-reported pre-pregnancy weight and standing height at the first study visit (<30 weeks gestation) measured by a study staff member via a stadiometer (Perspectives enterprises model PE-AIM-101). Pregnancy related covariates included, which was self-reported via interviewer- administered questionnaire at the first study visit, infant sex abstracted from EMR (96.8%), or self- reported from the participant (3.2%), and best available gestational age at time of birth for continuous birthweight models. In addition, information on maternal seafood consumption during the pregnancy was obtained via an interviewer-administered questionnaire in English or Spanish during the third trimester. Participants were asked if they had ever consumed any of the following types of seafood during the pregnancy: fish sticks, fresh oily fish, other fresh fish, canned tuna, shellfish, or fried shellfish. A combined variable was then created with 4 categories: never (62.6%), monthly (16.1%), at least weekly (12.3%), or unknown fish consumption status (9.0%). DAGitty was used to create a Directed Acyclic Graph (DAG) using the above covariates to assess relationships (Supplemental Figure 1) [274]. Minimal sufficient adjustment sets for estimating the total effect of prenatal PFAS exposure on birthweight were infant’s sex, country of birth, fish consumption, gestational age at time of blood sample, household income, maternal age, maternal education, parity, race/ethnicity and pre-pregnancy BMI. Continuous birthweight models were adjusted for infant’s sex (male, female), country of birth (US, other, missing indicator), fish consumption (never, monthly, at least weekly, missing indicator), gestational age at birth (weeks), 74 gestational age at time of blood sample (weeks), household income (<30k, 30-99k, >100k, or reported “Don’t Know”), maternal education (high school or less, some/completed college, or some graduate training), parity (first born, second or more born, missing indicator), race/ethnicity (Hispanic, non-Hispanic white, non-Hispanic Black or non-Hispanic other), pre-pregnancy BMI (kg/m 2 ), maternal age at study recruitment (years), and recruitment site. BW-for-GA z-score models included the same covariates except for gestational age at birth. Sex specific BW-for-GA z-score models removed infant sex while parity specific BW-for-GA z-score models removed parity. Statistical Analysis Univariate analyses were conducted with covariates to assess their relationships with the PFAS compounds ran continuously as well as with birthweight and gestational age at birth. Due to the right skewed distributions of the PFAS compounds, Kruskal-Wallis One Way Analysis of Variance (ANOVA) tests were used to assess relationships with categorical variables including education, household income, country of birth, race/ethnicity, parity, fish consumption and recruitment site. Spearman correlations were performed for associations between PFAS compounds and continuous variables including participant’s age, pre-pregnancy BMI, and gestational age at time of blood serum collection. Spearman correlations were also used to assess the correlation between the five compounds. For PFAS samples that were below the LOD, values were imputed using the equation of LOD/2 [275]. We used multiple linear regression models to assess the relationships between PFAS concentrations and birthweight and BW-for-GA z-scores that were parity or sex specific. Depending on the percentage of detected samples for each PFAS compound, we either modeled the exposure as a continuous exposure or categorical exposure (e.g., detected/non-detected or 75 quartiles of exposure). Because PFHxS and PFOS were detected in 100% of samples, we log transformed these exposures and modeled them on the continuous scale. We also modeled these two compounds categorically by quartiles. PFOA, PFNA, and PFDA were detected in 57-70% of samples; therefore we modeled these exposures categorically (detected versus non-detected). Additionally, multipollutant models were carried out that mutually adjusted for the five PFAS compounds. Single pollutant linear regression models and multipollutant models had no evidence of multicollinearity based on the variance inflation factor (VIF) (all individual variable VIF<10, mean VIF<2) and no influential points were identified (Cook’s D<1). Modeling assumptions for linear regression were checked and met for all models. A statistical interaction between each PFAS compound and infant sex was tested in single pollutant models and models were stratified by infant sex to assess for effect modification. A sensitivity analysis was carried out by excluding participants who had gestational diabetes mellitus (GDM) (N=33), gestational hypertension (N=26), or preeclampsia/eclampsia (N=25) recorded on their medical record, although some women overlapped with GDM and a gestational hypertension disorder (N=3). Lastly, an additional sensitivity analysis was carried out where models were restricted to only women who had prenatal PFAS concentrations measured in blood samples prior to 30 weeks gestation (N=239). Adjusted linear regression models were also performed to assess the relationship between prenatal PFAS blood serum and gestational age at birth to investigate whether gestational age was a biological intermediate between prenatal PFAS and infant birthweight [253, 276]. The significance level for all analyses was set at an alpha of 0.05 and all data were analyzed using SAS version 9.4 (SAS Institute Inc., Cary, NC, USA). 76 Results Participant Characteristics Participant and infant characteristics are reported in Table 1. Participants were on average 29 ± 6 years old at study recruitment, 76% were Hispanic, the majority (53%) had a high school diploma or less and had a mean pre-pregnancy BMI of 28 ± 6 kg/m 2 . Their infants were born at a mean of 39 ± 2 weeks gestation and weighed on average 3278 ± 522 grams at birth. The majority of infants were at least the second born child (62%). Maternal smoking during any point in pregnancy was low in this population (n = 6). Table 1. Demographics of 342 Mother-Infant Dyads Characteristic N (%) or Mean (SD) Maternal Age (years) 29.3 (6.0) Race/Ethnicity Hispanic Non-Hispanic Black Non-Hispanic White Non-Hispanic Other 260 (76.0%) 37 (10.8%) 30 (8.8%) 15 (4.4%) Country of Birth USA Other Unknown 175 (51.2%) 157 (45.9%) 10 (2.9%) Household Income <$50,000 $50,000-$99,999 >$100,000 Reported “Don’t Know” 165 (48.2%) 65 (19.0%) 28 (8.2%) 84 (24.6%) Education Completed high school or less Some college or completed college Some graduate training 181 (52.9%) 133 (38.9%) 28 (8.2%) Pre-Pregnancy BMI (kg/m 2 ) 28.2 (6.1) Any Personal Smoking During Pregnancy 6 (1.8%) Infant Gestational Age at Birth (weeks) 38.9 (1.8) Birthweight (grams) 3278.6 (522.2) Premature (<37 weeks) 36 (10.5%) Low Birth Weight (<2500 grams) 17 (5.0%) Male 176 (51.5%) Birth Order of Child First Born Second or more born Unknown 119 (34.8%) 212 (62.0%) 11 (3.2%) 77 PFAS Analytes Five PFAS analytes had greater than 50% of samples above the LOD (PFOS 100%, PFHxS 100%, PFNA 70%, PFOA 65%, and PFDA 57%). The remaining analytes had 17% or less of samples that were above the LOD and were excluded from this analysis. The median concentrations were highest for PFOS (1.34 ng/mL) followed by PFHxS (1.09 ng/mL). Distributions for all measured analytes are shown in Table 2. The natural log transformed PFAS concentrations were also significantly positively correlated with each other and are shown in Figure 2 (Spearman’s 𝜌 = 0.21 to 0.78, p value <0.0001 for all correlations). PFAS concentration levels seen within the MADRES cohort, with the exception of PFHxS, were lower than levels reported from comparable cohorts [277-279]. Table 2. Distribution of PFAS (ng/mL) Concentrations in Maternal Blood Serum (N=342) Analyte Abbreviation LOD (ng/mL) % Above LOD Min Q 1 Q2 Q 3 Max Perfluorooctanesulfonic acid PFOS 0.02 100% 0.09 0.97 1.34 1.86 10.36 Perfluorohexanesulfonic acid PFHxS 0.02 100% 0.36 0.79 1.09 1.47 4.1 Perfluorononanoic acid PFNA 0.02 70% ND ND 0.07 0.19 1.51 Perfluorooctanoic acid PFOA 0.035 65% ND ND 0.12 0.39 4.93 Perfluorodecanoic acid PFDA 0.035 57% ND ND 0.04 0.09 2.32 Perfluoro-n-pentanoic acid PFPEA 0.05 17% ND ND ND ND 1.19 Ethyl Perfluorooctane sulfonamido acetic acid NETFOSAA 0.02 12% ND ND ND ND 0.12 N-methyl Perfluorooctane sulfonamido acetic acid NMFOSAA 0.02 10% ND ND ND ND 0.38 Perfluroundecanoic acid PFUNDA 0.02 9% ND ND ND ND 0.72 Perfluorohexanoic acid PFHXA 0.05 1% ND ND ND ND 0.11 Perfluorobutanesulfonic acid PFBS 0.02 1% ND ND ND ND 0.08 Perfluorododecanoic acid PFDODA 0.035 1% ND ND ND ND 0.94 Perfluorooctanesulfonamide PFOSA 0.02 0% ND ND ND ND ND Perfluoroheptanoic acid PFHPA 0.05 0% ND ND ND ND ND Note: ND=Not Detected Q1=Quartile 1 (25 th percentile) Q2= Quartile 2 (50 th percentile) Q3=Quartile 3 (75 th percentile) 78 Figure 2. Spearman Correlation Coefficients for Natural Log Transformed PFAS (ng/mL) in Maternal Blood Serum (N = 342) Note: p<0.0001 for all correlations Median levels of PFAS by demographic characteristics are shown in Supplemental Table 1. Significant differences were observed in median PFOA, PFNA, PFHxS, and PFOS by education level and participants with at least some graduate training had the highest levels. These trends were consistent with household income with median PFOA, PFNA, PFHxS, and PFOS differing across income bracket and those making at least $100,000 annually having the highest levels. Significant differences in median PFOA, PFNA, PFHxS, and PFOS were observed by maternal race with non- Hispanic white participants having the highest levels. Additionally, there were significant differences in median PFOA, PFDA, PFNA, PFHxS, and PFOS by parity status with participants pregnant with their first child having the highest levels. Median PFNA, PFHxS, and PFOS significantly differed by fish consumption habits during pregnancy with those who consumed fish at least weekly having the highest levels. Significant differences in median PFOS and PFHxS were observed by country of birth, with participants born outside of the US having lower levels. All PFAS concentrations were significantly inversely associated with pre-pregnancy BMI 79 (Spearman’s 𝜌 = -0.29 to -0.13, p value <0.02 for all correlations) while no relationships were observed with PFAS and maternal age (p values >0.14 for all correlations). Single Pollutant Linear Regression Models No significant associations were observed with any PFAS compound and gestational age at birth in univariate or adjusted models (Supplemental Table 2). Therefore, we concluded gestational age was not acting as a mediator of the association between prenatal PFAS exposure and birthweight within the MADRES cohort. We included gestational age at birth as a covariate in continuous birthweight models. Results for single prenatal PFAS exposures with continuous infant birthweight and BW-for-GA z-scores that were parity or sex specific are shown in Table 3. Women with detected levels of PFOA (𝛽 = -119.7 grams; 95% CI -216.7, -22.7) and PFNA (𝛽 = -105.8 grams; 95% CI -206.3, -5.3) had infants with significantly lower birthweight on average compared to women without detected levels. No significant associations were seen with the remaining PFAS compounds and infant birthweight, although non-significant inverse relationships were seen with PFOS and PFDA and a non-significant positive relationship was seen with PFHxS. Results for all PFAS were similar in BW-for-GA z-score models compared to continuous birthweight models, although only PFOA remained significant in models with sex specific BW- for-GA z-scores (𝛽 = -0.3; 95% CI -0.5, -0.1) and parity specific BW-for-GA z-scores (𝛽 = -0.3; 95% CI -0.5, -0.1). No evidence was found for a statistical interaction between any PFAS compound and infant sex (all interaction p values >0.29) and there were no significant differences in stratified models by sex (data not shown). 80 Table 3. Adjusted Single PFAS Exposure Models with Continuous Birthweight and Sex or Parity Specific BW-for-GA Z-Scores (N=342) Model Birthweight 1 𝛽 (95% CI) BW-for-GA z-scores, sex 2 𝛽 (95% CI) BW-for-GA z-scores, parity 3 𝛽 (95% CI) PFOS <0.97 ng/mL 0.97-1.33 ng/mL 1.34-1.85 ng/mL ≥1.86 ng/mL -37.2 § (-123.9, 49.6) REF -41.5 (-167.9, 85.0) -51.4 (-178.0, 75.2) -53.9 (-188.8, 80.9) -0.1 § (-0.3, 0.1) REF -0.1 (-0.5, 0.2) -0.1 (-0.4, 0.2) -0.1 (0.5, 0.2) -0.1 § (-0.3, 0.2) REF -0.2 (-0.5, 0.2) -0.1 (-0.4, 0.2) -0.1 (-0.5, 0.2) PFHxS < 0.79 ng/mL 0.79- 1.08 ng/mL 1.09-1.46 ng/mL ≥1.47 ng/mL 49.7 § (-78.7, 178.2) REF 81.6 (-44.1, 207.4) 24.4 (-109.6, 158.3) 118.1 (-32.3, 267.5) 0.1 § (-0.2, 0.5) REF 0.2 (-0.1, 0.7) 0.1 (-0.3, 0.4) 0.3 (-0.1, 0.7) 0. 2 § (-0.2, 0.5) REF 0.2 (-0.1, 0.6) 0.1 (-0.2, 0.4) 0.3 (-0.1, 0.7) PFOA ≤LOD >LOD REF -119.7 (-216.7, -22.7)* REF -0.3 (-0.5, -0.1)* REF -0.3 (-0.5, -0.1)* PFNA ≤LOD >LOD REF -105.8 (-206.3, -5.3)* REF -0.2 (-0.5, 0.01) REF -0.2 (-0.5, 0.01) PFDA ≤LOD >LOD REF -55.4 (-145.1, 34.3) REF -0.2 (-0.4, 0.1) REF -0.2 (-0.4, 0.1) Note: CI=confidence interval; PFHxS=Perfluorohexanesulfonic acid; PFOS=Perfluorooctanesulfonic acid; PFNA=Perfluorononanoic acid; PFOA=Perfluorooctanoic acid; PFDA=Perfluorodecanoic acid. Reference group: Samples <LOD; PFNA LOD=0.02 ng/mL, PFDA LOD=0.035 ng/mL, PFOA LOD=0.035 ng/mL. 1 Adjusted for maternal race, maternal age at study recruitment, country of birth, maternal education, maternal household income, recruitment site, GA at time of blood sample, pre-pregnancy BMI, pregnancy fish consumption, GA at birth, parity and infant sex 2 Adjusted for maternal race, maternal age at study recruitment, country of birth, maternal education, maternal household income, recruitment site, GA at time of blood sample, pre-pregnancy BMI, pregnancy fish consumption, and parity 3 Adjusted for maternal race, maternal age at study recruitment, country of birth, maternal education, maternal household income, recruitment site, GA at time of blood sample, pre-pregnancy BMI, pregnancy fish consumption, and infant sex, N=331 § Beta estimate interpreted as per natural log increase *p<0.05 Multipollutant Linear Regression Models In models that mutually adjusted for all five PFAS compounds the effect estimate was slightly attenuated from independent models for detected/non-detected PFOA and continuous birthweight and detected/non-detected PFOA was no longer statistically significant (𝛽 = -108.8 grams; 95% CI -229.9, 12.3). In contrast, the effect estimates for detected/non-detected PFOA and parity specific BW-for-GA z-scores and sex specific BW-for-GA z-scores remained unchanged 81 from single pollutant to multipollutant models, although statistical significance was only reached in the parity specific model (𝛽 = -0.3; 95% CI -0.6, -0.01). PFNA, PFDA, PFOS, and PFHxS were not associated with any birthweight measure in multipollutant models. Results are shown in Table 4. Table 4. Linear Regression Model Results for 5 PFAS Compounds Simultaneously on Continuous Infant Birthweight and Sex or Parity Specific BW-for-GA Z-Scores (N=342) PFAS Birthweight 1 𝛽 (95% CI) BW-for-GA z-scores, sex 2 𝛽 (95% CI) BW-for-GA z-scores, parity 3 𝛽 (95% CI) PFOS PFHxS PFOA (≤LOD) PFOA (> LOD) PFNA (≤LOD) PFNA (>LOD) PFDA (≤LOD) PFDA (>LOD) 29.3 § (-81.2, 139.7) 85.6 § (-48.9, 220.1) REF -108.8 (-229.9, 12.3) REF -57.9 (-185.9, 69.9) REF -39.2 (-138.9, 60.4) 0.1 § (-0.2, 0.3) 0.2 § (-0.1, 0.6) REF -0.3 (-0.6, 0.02) REF -0.1 (-0.4, 0.2) REF -0.1 (-0.4, 0.1) 0.1 § (-0.2, 0.4) 0.2 § (-0.1, 0.6) REF -0.3 (-0.6, -0.01)* REF -0.1 (-0.4, 0.2) REF -0.2 (-0.4, 0.1) Note: CI=confidence interval; PFHxS=Perfluorohexanesulfonic acid; PFOS=Perfluorooctanesulfonic acid; PFNA=Perfluorononanoic acid; PFOA=Perfluorooctanoic acid; PFDA=Perfluorodecanoic acid. Reference group: Samples <LOD; PFNA LOD=0.02 ng/mL, PFDA LOD=0.035 ng/mL, PFOA LOD=0.035 ng/mL 1 Adjusted for maternal age, maternal race/ethnicity, maternal education, country of birth, maternal household income, recruitment site, GA at time of blood sample, pre-pregnancy BMI, pregnancy fish consumption, infant sex, GA at birth, parity 2 Adjusted for maternal age, maternal race/ethnicity, maternal education, country of birth, maternal household income, recruitment site, GA at time of blood sample, pre-pregnancy BMI, pregnancy fish consumption, and parity 3 Adjusted for maternal age, maternal race/ethnicity, maternal education, country of birth, maternal household income, recruitment site, GA at time of blood sample, pre-pregnancy BMI, pregnancy fish consumption, and infant sex, N=331 § Beta estimate interpreted as per natural log increase *p<0.05 Sensitivity Analyses In models that were restricted to participants without GDM, gestational hypertension or preeclampsia/eclampsia recorded on their medical records (N=261), estimates were comparable to both single pollutant and multipollutant models with the full sample (Supplemental Table 3). We also found that restricting to women with PFAS concentrations measured prior to 30 weeks gestation (N=239) did not materially change our results (Supplemental Table 4). Discussion We found that women with detectable levels of PFOA during pregnancy had infants that weighed on average 120 grams less than women without detectable levels of PFOA. In models 82 mutually adjusting for five detected PFAS compounds, the effect size for the associations of birthweight with PFOA remained comparable. We did not find significant adjusted associations with the remaining PFAS concentrations and the birth outcomes assessed. Additionally, many PFAS compounds analyzed within the MADRES cohort had limited to no detection. In the MADRES cohort, median PFNA, PFOA, PFOS, and PFDA concentrations were lower in relation to several other pregnancy cohorts within Asian, European, and North American countries, although this trend was not consistent with PFHxS, which was often comparable or had higher concentration levels [263, 264, 277-279]. Although we do not know why PFHxS concentrations were higher in this cohort, PFHxS does have one of the longest half-lives of PFAS compounds (5-9 years) [233, 280]. The highest levels of PFAS concentrations were found in participants who were pregnant with their first-born child. This is consistent with prior literature [281-286], as breastfeeding has been shown to be a probable excretion route for PFAS [283, 285]. Prior studies have also reported higher levels of PFAS within non-Hispanic white women, although research is scarce for comparisons to Hispanic women outside of NHANES. PFOA and PFHxS have been seen to be higher in non-Hispanic white women compared to Black/African- American women [287] as well as to Chinese-American and Japanese-American women [288]. These results are surprising given that communities of color are disproportionately burdened by environmental chemical exposures including hazardous waste sites, air pollution, and water pollution [289]. PFAS exposure sources related to race and ethnicity merit further examination. The observation that higher education levels and annual household income were associated with higher PFAS levels within MADRES is also consistent with previous studies [282, 290, 291]. Foreign-born participants having lower levels of PFOS and PFHxS as seen in this study is consistent with a previous study of middle-aged women, which found that within Chinese and 83 Japanese women, those born outside of the US had lower PFAS concentrations compared to those born within the US [288]. Our findings that PFAS levels were inversely related with pre-pregnancy BMI are more complex. A previous study in Norway consisting of white and primarily of normal weight women, showed prenatal PFHxS levels to be significantly positively associated with pre-pregnancy BMI [290], which is contrary to our findings. We found that PFAS levels were higher among women who reported higher intakes of seafood. This is consistent with a study among Norwegian women in which unadjusted PFAS serum levels were associated with increased dietary intakes of shellfish and oily fish [290]. The majority of studies on prenatal PFAS exposure and infant birthweight have investigated individual PFAS exposures. Steenland et al., found prenatal PFOA measured in maternal or cord blood to be associated with a decrease of 10.5 grams (95% CI -16.7, -4.4) in infant birthweight per ng/mL increase in PFOA in the primary random effects meta-analysis [255]. Starling et al. found similar significant results to those reported in this current study in 628 mother- infant pairs within the Healthy Start Study. They found that women in the highest tertile of mid- pregnancy measured PFOA exposure had infants that weighed significantly less compared to those in the lowest tertile of PFOA exposure (𝛽 = -92.4 grams; 95% CI -166.2, -18.5) [279]. Another recent study found prenatal PFOA measured in first trimester serum to be significantly associated with decreased infant birthweight in 1,533 participants from the Swedish Environmental, Longitudinal, Mother and child, Asthma and allergy (SELMA) study. Mothers in the highest quartile of PFOA exposure had infants that weighed 90 grams less (95% CI -159, -91) compared to those in the lowest quartile [259]. 84 In this present study, we observed no consistent significant findings between PFDA, PFNA, PFOS, and PFHxS and infant birthweight. However, we found non-significant inverse associations for prenatal PFNA, PFDA, and PFOS with birthweight and non-significant positive association was observed with PFHxS and birthweight. Previous studies have shown significant inverse relationships between PFOS exposure and birthweight in a subset of several studies [292- 296] and three other studies have found non-significant positive relationships with PFHxS and birthweight [257, 260, 294]. Few studies have examined relationships between PFNA and PFDA with birthweight, although significant inverse relationships have been observed [257, 259, 261, 262, 264]. Our study did not find any significant results with PFAS and gestational age at birth and therefore gestational age was not considered to be a likely mediator of the association of PFAS on infant birthweight [253, 276]. Several other studies have also not found significant associations between prenatal PFAS exposure and gestational age at birth [265, 277, 297]. Many important strengths are found within the current study. First, multiple PFAS exposures were analyzed with the same panel and, in particular, we had measurements of PFDA, PFNA, and PFHxS, which have not been as comprehensively studied as PFOA and PFOS. Second, PFAS concentrations were measured in maternal blood serum, which is a superior biospecimen to whole blood or urine, with the serum to whole blood ratio for PFAS being 2:1 [298, 299]. Third, our study population is comprised of predominately low-income Hispanic women, which is an understudied group. Additionally, given pregnant persons are exposed to more than one compound at once, we also assessed associations within a multipollutant framework and found that PFOA was consistently associated with birthweight. Lastly, the prospective cohort design and high- quality covariate data are additional strengths. 85 Limitations of this study include only one measurement of PFAS concentrations and PFAS were measured in serum collected across a wide range of gestational age. For these reasons, this analysis was unable to determine critical windows of exposure of PFAS on infant birthweight. As such, while we did control for gestational age at the time of collection, we assume that PFAS levels are relatively stable across pregnancy due to their long half-lives. A previous study with PFOA and PFOS blood serum levels taken from the first and second trimester found a high degree of correlation between the two time points, but the averages were lower in the second trimester [293]. An additional study with blood samples taken at all three trimesters of pregnancy found PFOA, PFOS, PFNA, and PFDA to slightly decrease over pregnancy while PFHxS levels remained consistent [257]. We also did not have available information on glomerular filtration rate (GFR), which can influence both PFAS levels later in pregnancy and infant birthweight [300] as GFR peaks during 30-35 weeks gestation [301]. In our study, we conducted a sensitivity analysis limiting our analysis to samples collected prior to 30 weeks gestation and we found no measurable impact on our effect estimates. Additionally, although many key covariates were adjusted for in final models, residual confounding is still possible, as in all observational studies. Conclusion Prenatal exposure to PFOA was associated with lower birthweight in infants in the MADRES cohort in Los Angeles, CA. These findings suggest that exposure to these chemicals during in utero development may have important implications for children’s health. While we found the highest exposures to PFAS among non-Hispanic white women as well as women with higher household incomes and higher attained education levels in our cohort, future studies should focus on how mixtures of newer PFAS, especially those that have become substitutes for legacy 86 PFAS including PFOA and PFOS, impact populations facing disparities in birth and other health outcomes. 87 Supplemental Material Supplemental Figure 1. Directed Acyclic Graph (DAG) of PFAS Exposure and Infant Birthweight Note: Diagram was visualized using DAGitty software. 88 Supplemental Table 1. Median PFAS Analyte Concentrations (ng/mL) by Maternal Demographic Characteristics PFOS PFHxS PFOA PFDA PFNA Race/Ethnicity Hispanic (n=260) Non-Hispanic Black (n=37) Non-Hispanic White (n=30) Non-Hispanic Other (n=15) 1.27 ** 1.45 ** 1.95 ** 1.76 ** 1.01 ** 1.08 ** 1.81 ** 1.34 ** 0.11 ** 0.05 ** 0.52 ** 0.35 ** 0.04 0.04 0.05 0.08 0.07 * 0.06 * 0.16 * 0.14 * Household Income <$50,000 (n=165) $50,000-$99,999 (n=65) >$100,000 (n=28) Reported “Don’t Know” (n=84) 1.27 ** 1.48 ** 1.99 ** 1.26 ** 1.05 ** 0.98 ** 1.69 ** 1.13 ** 0.10 ** 0.19 ** 0.59 ** 0.06 ** 0.05 0.04 0.07 0.04 0.06 ** 0.11 ** 0.23 ** 0.06 ** Education Completed high school or less (n=181) Some college of completed college (n=133) Some graduate training (n=28) 1.26 ** 1.38 ** 2.06 ** 1.00 ** 1.15 ** 1.49 ** 0.08 ** 0.12 ** 0.53 ** 0.04 0.04 0.05 0.06 ** 0.09 ** 0.21 ** Country of Birth USA (n=175) Other (n=157) Unknown (n=10) 1.44 * 1.26 * 1.08 * 1.13 * 1.01 * 1.30 * 1.18 0.09 0.10 0.04 0.05 0.03 0.09 0.07 0.06 Birth Order of Child First Born (n=119) Second or more born (n=212) Unknown (n=11) 1.63 ** 1.25 ** 1.21 ** 1.25 ** 0.99 ** 1.30 ** 0.35 ** 0.07 ** 0.10 ** 0.05 * 0.04 * 0.05 * 0.16 ** 0.06 ** 0.07 ** Pregnancy Fish Consumption Never (n=214) Monthly (n=55) At least Weekly (n=42) Unknown (n=31) 1.26 * 1.43 * 1.73 * 1.38 * 1.07 * 1.19 * 1.28 * 0.90 * 0.09 0.18 0.23 0.22 0.04 0.04 0.07 0.02 0.06 * 0.11 * 0.14 * 0.07 * *Kruskal-Wallis One Way ANOVA p<0.05 **Kruskal-Wallis One Way ANOVA p<0.0001 89 Supplemental Table 2. Univariate and Adjusted Models for Independent Prenatal PFAS Compounds and Gestational Age at Birth, in Weeks (N=342) Compound Unadjusted 𝛽 (95% CI) Adjusted* 𝛽 (95% CI) PFHxS < 0.79 ng/mL 0.79- 1.08 ng/mL 1.09-1.46 ng/mL ≥1.47 ng/mL 0.3 1 (-0.1, 0.8) REF -0.1 (-0.6, 0.5) -0.1 (-0.6, 0.5) 0.4 (-0.2, 0.9) 0.1 1 (-0.5, 0.6) REF -0.03 (-0.5, 0.5) -0.3 (-0.8, 0.3) 0.2 (-0.5, 0.5) PFOS <0.97 ng/mL 0.97-1.33 ng/mL 1.34-1.85 ng/mL ≥1.86 ng/mL 0.2 1 (-0.2, 0.5) REF 0.3 (-0.3, 0.8) 0.003 (-0.5, 0.5) 0.1 (-0.4, 0.7) -0.3 1 (-0.4, 0.3) REF 0.2 (-0.3, 0.7) -0.1 (-0.6, 0.4) -0.2 (-0.7, 0.4) PFDA ≤LOD >LOD REF 0.1 (-0.3, 0.5) REF 0.02 (-0.3, 0.4) PFNA ≤LOD >LOD REF -0.1 (-0.5, 0.3) REF -0.3 (-0.7, 0.1) PFOA ≤LOD >LOD REF 0.002 (-0.4, 0.4) REF -0.1 (-0.5, 0.3) Note: CI=confidence interval; PFHxS=Perfluorohexanesulfonic acid; PFOS=Perfluorooctanesulfonic acid; PFNA=Perfluorononanoic acid; PFOA=Perfluorooctanoic acid; PFDA=Perfluorodecanoic acid. Reference group: Samples <LOD PFNA LOD=0.02 ng/mL, PFDA LOD=0.035 ng/mL, PFOA LOD=0.035 ng/mL. * Adjusted for maternal age, pre-pregnancy BMI, maternal education, country of birth, household income, mother’s race, gestational age at time of blood sample, study recruitment site, parity, infant sex, and fish consumption 1 Beta estimate represents a per natural log increase 90 Supplemental Table 3. Maternal Serum PFAS and Birthweight Models Restricted to Participants Without Preeclampsia/Eclampsia, Gestational Hypertension or Gestational Diabetes Mellitus (GDM) (N=261) Model Birthweight 1 𝛽 (95% CI) BW-for-GA, sex 2 𝛽 (95% CI) BW-for-GA, parity 3 𝛽 (95% CI) Single Pollutant Models PFOS <0.97 ng/mL 0.97-1.33 ng/mL 1.34-1.85 ng/mL ≥1.86 ng/mL -17.5 § (-120.3, 85.2) REF -41.5(-187.9, 104.9) -24.4 (-169.8, 120.9) -22.2 (-176.7, 132.3) -0.1 § (-0.3, 0.2) REF -0.1 (-0.5, 0.2) -0.1 (-0.5, 0.3) -0.1 (-0.5, 0.3) -0.1 § (-0.3, 0.2) REF -0.2 § (-0.6, 0.2) -0.1 (-0.5, 0.3) -0.2 (-05, 0.2) PFHxS < 0.79 ng/mL 0.79- 1.08 ng/mL 1.09-1.46 ng/mL ≥1.47 ng/mL 106.6 § (-48.1, 261.2) REF 93.8 (-50.4, 238.0) 77.4 (-85.5, 240.4) 190.7 (14.6, 366.8)* 0.3 § (-0.1, 0.7) REF 0.3 (-0.1, 0.6) 0.2 (-0.2, 0.6) 0.5 (0.02, 0.9)* 0.3 § (-0.1, 0.7) REF 0.3 (-0.1, 0.6) 0.2 (-0.2, 0.6) 0.5 (0.01, 0.9)* PFOA ≤LOD >LOD REF -121.9 (-232.9, -10.9)* REF -0.3 (-0.6, -0.03)* REF -0.3 (-0.6, -0.1)* PFNA ≤LOD >LOD REF -114.7 (-229.2, -0.2)* REF -0.3 (-0.6, 0.003) REF -0.3 (-0.6, -0.01)* PFDA ≤LOD >LOD REF -34.9 (-138.2, 68.3) REF -0.1 (-0.4, 0.1) REF -0.2 (-0.4, 0.1) Multipollutant Model PFOS PFHxS PFOA (≤LOD) PFOA (> LOD) PFNA (≤LOD) PFNA (>LOD) PFDA (≤LOD) PFDA (>LOD) 47.3 § (-78.5, 173.0) 126.3 § (-32.3, 284.9) REF -107.2 (-247.8, 33.4) REF -81.8 (-230.8, 67.3) REF -21.9 (-135.6, 91.8) 0.1 § (-0.2, 0.4) 0.3 § (-0.1, 0.7) REF -0.3 (-0.6, 0.1) REF -0.2 (-0.5, 0.2) REF -0.1 (-0.4, 0.2) 0.1 § (-0.2, 0.5) 0.3 § (-0.1, 0.7) REF -0.3 (-0.7, 0.02) REF -0.2 (-0.6, 0.2) REF -0.1 (-0.4, 0.2) Note: CI= confidence interval; PFHxS=Perfluorohexanesulfonic acid; PFOS=Perfluorooctanesulfonic acid; PFNA=Perfluorononanoic acid; PFOA=Perfluorooctanoic acid; PFDA=Perfluorodecanoic acid. Reference group: Samples <LOD PFNA LOD=0.02 ng/mL, PFDA LOD=0.035 ng/mL, PFOA LOD=0.035 ng/mL. Multipollutant model is mutually adjusted for all five analytes. 1 Adjusted for maternal race, maternal age at study recruitment, maternal education, country of birth, maternal household income, recruitment site, gestational age at time of blood sample, pre-pregnancy BMI, pregnancy fish consumption, infant sex, gestational age at birth, parity 2 Adjusted for maternal race, maternal age at study recruitment, maternal education, country of birth, maternal household income, recruitment site, gestational age at time of blood sample, pre-pregnancy BMI, pregnancy fish consumption, and parity 3 Adjusted for maternal race, maternal age at study recruitment, maternal education, country of birth, maternal household income, recruitment site, gestational age at time of blood sample, pre-pregnancy BMI, pregnancy fish consumption, and infant sex; N=251 *p<0.05 § Beta estimate interpreted as per natural log increase 91 Supplemental Table 4. Maternal Serum PFAS and Birthweight Models Restricted to Samples Collected Prior to 30 Weeks Gestation (N=239) Model Birthweight 1 𝛽 (95% CI) BW-for-GA, sex 2 𝛽 (95% CI) BW-for- GA, parity 3 𝛽 (95% CI) Single Pollutant Models PFOS <0.97 ng/mL 0.97-1.33 ng/mL 1.34-1.85 ng/mL ≥1.86 ng/mL -53.3 § (-166.9, 60.2) REF -11.1 (-172.4, 150.2) -48.4 (-212.0, 115.2) -60.2 (-227.5, 107.1) -0.1 § (-0.4, 0.1) REF -0.06 (-0.5, 0.3) -0.1 (-0.5, 0.3) -0.2 (-0.6, 0.3) -0.2 § (-0.4, 0.1) REF -0.1 (-0.5, 0.3) -0.1 (-0.5, 0.3) -0.2 (-0.6, 0.2) PFHxS < 0.79 ng/mL 0.79- 1.08 ng/mL 1.09-1.46 ng/mL ≥1.47 ng/mL 36.7 § (-125.3, 198.8) REF 71.7 (-71.7, 215.0) 15.5 (-145.5, 176.5) 119.7 (-81.3, 320.6) 0.1 § (-0.3, 0.5) REF 0.2 (-0.2, 0.5) 0.04 (-0.4, 0.4) 0.3 (-0.2, 0.8) 0.1 § (-0.3, 0.5) REF 0.2 (-0.1, 0.6) 0.1 (-0.3, 0.5) 0.3 (-0.2, 0.8) PFOA ≤LOD >LOD REF -121.0 (-242.4, 0.3) REF -0.3 (-0.6, -0.003)* REF -0.3 (-0.6, -0.007)* PFNA ≤LOD >LOD REF -127.7 (-254.2, -1.2)* REF -0.3 (-0.6, -0.01)* REF -0.3 (-0.7, -0.03)* PFDA <=LOD >LOD REF -85.9 (-196.4, 24.6) REF -0.2 (-0.5, 0.05) REF -0.3 (-0.5, 0.02) Multipollutant Model PFOS PFHxS PFOA (≤LOD) PFOA (> LOD) PFNA (≤LOD) PFNA (>LOD) PFDA (≤LOD) PFDA (>LOD) 44.5 § (-110.1, 199.1) 69.3 § (-103.1, 241.6) REF -97.5 (-258.2, 63.3) REF -77.9 (-245.3, 89.4) REF -67.4 (-192.0, 57.3) 0.1 § (-0.3, 0.5) 0.2 § (-0.2, 0.6) REF -0.2 (-0.6, 0.2) REF -0.2 (-0.6, 0.2) REF -0.2 (-0.5, 0.1) 0.1 § (-0.3, 0.5) 0.2 § (-0.2, 0.7) REF -0.2 (-0.6, 0.2) REF -0.2 (-0.5, 0.1) REF -2 (-0.5, 0.1) Note: CI= confidence interval; PFHxS=Perfluorohexanesulfonic acid; PFOS=Perfluorooctanesulfonic acid; PFNA=Perfluorononanoic acid, PFOA = Perfluorooctanoic acid; PFDA=Perfluorodecanoic acid. Reference group: Samples <LOD PFNA LOD=0.02 ng/mL, PFDA LOD=0.035 ng/mL, PFOA LOD=0.035 ng/mL. Multipollutant model is mutually adjusted for all five analytes. 1 Adjusted for maternal race, maternal age at study recruitment, maternal education, country of birth, maternal household income, recruitment site, gestational age at time of blood sample, pre-pregnancy BMI, pregnancy fish consumption, infant sex, gestational age at birth, parity 2 Adjusted for maternal race, maternal age at study recruitment, maternal education, country of birth, maternal household income, recruitment site, gestational age at time of blood sample, pre-pregnancy BMI, pregnancy fish consumption, and parity 3 Adjusted for maternal race, maternal age at study recruitment, maternal education, country of birth, maternal household income, recruitment site, gestational age at time of blood sample, pre-pregnancy BMI, pregnancy fish consumption, and infant sex *p<0.05 § Beta estimate interpreted as per natural log increase 92 CHAPTER FIVE Study 2: Prenatal Perfluorooctanoic Acid (PFOA) Exposure Adversely Impacted Fetal Head Biometric Parameters in Participants Experiencing Higher Perceived Stress During Pregnancy Within the MADRES Cohort Abstract Introduction Perfluoroalkyl substances (PFAS) are ubiquitous synthetic chemicals with long half-lives and are known to cross the placenta during pregnancy. We examined the influence of maternal PFAS levels on in utero fetal growth trajectories in a structurally marginalized population and assessed whether maternal stress modified these associations. Methods Blood serum concentrations of 14 PFAS were measured in 335 prenatal specimens (mean gestational age (GA): 21±9 weeks) in the ongoing Maternal And Developmental Risks from Environmental and Social Stressors (MADRES) cohort. This study included analytes (PFOS, PFHxS, PFNA, PFOA, PFDA) detected in >50% of samples. Fetal growth outcomes were abstracted from ultrasound medical records, and head circumference, abdominal circumference, biparietal diameter, femur length, and estimated fetal weight were also measured during the 3 rd trimester study visit by certified sonographers (N=833 scans, GA range 10-42 weeks, mean 2.4 scans/participant). We estimated the association between each prenatal PFAS exposure and fetal growth trajectory by fitting a linear mixed effects model that allowed for a quadratic growth curve in GA, with adjustment for key covariates. PFOS and PFHxS were modeled continuously (100% 93 sample detection), while PFOA, PFNA, and PFDA were modeled categorically (detected vs. non- detected) due to low sample detection (57%-70%). Scores on the Perceived Stress Scale (PSS) measured in pregnancy were dichotomized at the sample median (<13 vs. ≥13) in stratified models. Results Participants were on average 29±6 years old at study entry and were predominately Hispanic (76%). Median serum concentrations of PFOS, PFHxS, PFNA, PFOA and PFDA were 1.34, 1.10, 0.07, 0.12, and 0.04 ng/mL, respectively. Participants with detected PFOA concentrations had fetuses with -2.5 mm (95% CI -4.2, -0.8) smaller head circumference on average than those without detected PFOA concentrations. Additionally, participants with detected PFOA concentrations had fetuses with -0.7 mm (95% CI -1.3, -0.2) smaller biparietal diameter on average than those without detected PFOA concentrations. In models stratified by PSS level, the effects of PFOA on fetal growth parameters were stronger and only significant in participants with higher perceived stress (Head Circumference: 𝛽 = -3.5, 95% CI -5.8, -1.4; Biparietal Diameter: 𝛽 = -0.8, 95% CI -1.6, - 1.1). Conclusion In utero exposure to higher PFOA concentrations was associated with lower fetal head biometry parameters among participants with higher perceived stress, suggesting that the combination of PFAS exposure and stress may negatively impact fetal growth. 94 Introduction Per –and polyfluoroalkyl Substances (PFAS) are a class of synthetic chemicals that consists of over 4,700 separate compounds [302]. Due to their oil and water-resistant qualities, they have been commonly used in a wide range of items including non-stick cookware, shampoos, waterproof clothing, fire-fighting foam and fast-food packaging [302-304]. The strong and stable bond created between carbon and fluorine [305] make these compounds persistent in the environment and the most common exposure route to humans is through ingestion (i.e. dietary fish intake and drinking water) [306]. Given their association with numerous health problems [307- 310] perfluorooctane sulfonic acid (PFOS) and perfluorooctanoic acid (PFOA) were phased out of production in the United States in the early 2000s [311]. However, due to the long half-lives of PFAS [312], these compounds as well as other PFAS have been found in >99% of blood samples from U.S adults and children participating in the National Health and Nutrition Examination Survey (NHANES) study population in recent years [313, 314]. A particularly susceptible period of exposure to PFAS may be during pregnancy due to the biological changes occurring in the mother [315] as well as the ability for PFAS to cross the placental barrier to reach the fetus and bioaccumulate in the placenta [316, 317]. Implications of PFAS exposure for fetal growth are important to understand, as fetal development may predict health outcomes across the life course. Low birthweight (<2,500 grams) increases risk for metabolic health conditions later in life [318-320] and in utero growth parameters are significantly associated with childhood cognitive outcomes and obesity risk [321-323]. Substantial literature has indicated that exposure to PFAS, particularly PFOA and PFOS, during pregnancy is associated with lower infant birthweight [324-327], however, there are few studies that have explored the influence on growth in utero by standard ultrasound biometry measurements of fetal head 95 circumference, biparietal diameter, abdominal circumference, femur length, and estimated fetal weight [328, 329]. Pregnant persons of color, notably non-Hispanic Black and Hispanic populations, have higher rates of low birthweight infants compared to their non-Hispanic white counterparts [330]. In some studies, PFAS levels have also differed by racial and ethnic groups. A study conducted within NHANES found that PFOS levels were highest in non-Hispanic Black persons, followed by non-Hispanic white persons, and lowest in Mexican-American individuals [313]. In contrast, a small cohort of middle-aged women (N=178) found levels of PFOA and perfluorohexanesulfonic acid (PFHxS) to be higher in non-Hispanic white women compared to Black women. A United States cohort comprised of 12 clinical sites assessed a panel of 11 PFAS analytes and found that pregnant persons who were non-Hispanic white or Asian tended to have higher levels for most analytes compared to Black and Hispanic pregnant persons [329]. However, studies assessing PFAS exposure on fetal growth in health disparities populations and those specifically in Hispanic/Latina populations are currently limited. Health disparity populations face unique stressors related to structural inequities based on race and ethnicity, discrimination, low-income levels, and negative stereotypes [331]. It is well documented that psychological distress such as perceived stress experienced prior to pregnancy and during pregnancy influences adverse pregnancy outcomes including low infant birthweight [332, 333]. Previous research has suggested that multiple exposures to chemicals and psychosocial stress may exhibit a cumulative effect on fetal growth [334]. The objective of this study was to assess the influence of five PFAS analytes including PFOS, PFOA, PFHxS, perfluorononanoic acid (PFNA) and perfluorodecanoic acid (PFDA) from a panel of fourteen PFAS analytes quantified during pregnancy on longitudinal fetal growth 96 trajectories of fetal head circumference, biparietal diameter, abdominal circumference, femur length, and estimated fetal weight within a structurally marginalized population. Additionally, this study explored whether perceived stress experienced during pregnancy modified these associations. We hypothesized that participants with higher prenatal PFAS exposure would have infants with decreased fetal growth trajectories and that pregnant persons with higher stress would be more vulnerable to the health effects of PFAS. Methods Study Sample This study included a subset of participants from the ongoing Maternal And Developmental Risks from Environmental and Social Stressors (MADRES) pregnancy cohort. Details related to the study design, protocol, and demographics of the cohort have previously been described [335]. Briefly, participants were eligible if they were less than 30 weeks pregnant, at least 18 years of age, and a fluent speaker of English or Spanish. Exclusion criteria for the study included multiple gestation, the inability to participate and provide consent due to a physical, mental or cognitive disability, current incarceration, or HIV positive status. Recruitment occurred during pregnancy from four prenatal clinic sites including two community health centers, one county hospital prenatal clinic, and one private obstetrics and gynecology practice. At time of study enrollment, informed consent and HIPAA authorization to access medical records were obtained from each participant and the University of Southern California’s Institutional Review Board (IRB) approved all aspects of the study. Participants from the larger MADRES cohort were included in the current study if the following criteria were met: 1) data were available for PFAS concentrations measured in blood serum, 2) at least one prenatal ultrasound scan with fetal biometry measurements was available, 97 and 3) data on maternal race/ethnicity and fetal sex were present. PFAS concentrations were measured in Spring 2019 in blood serum samples that were collected from 359 participants from December 2015 until February 2019. Of these 359 participants, 344 had at least one prenatal ultrasound (859 total scans). We removed five participants with missing information on race/ethnicity and four participants with missing data on fetal sex resulting in a final sample size of 335 unique participants with a total of 833 ultrasound scans. The consort diagram illustrating those included is shown in Figure 1. Figure 1. Consort Diagram of Included Participants Serum PFAS Analysis Maternal blood samples were collected at an in-person study visit (median gestational age 19.0 weeks, range 5.7-38.3 weeks) using red top 10-milliliter serum tubes. The blood samples were processed, and serum was aliquoted and then stored at -80°C until shipment to the Wadsworth Center’s Human Health Exposure Assessment Resource (WC-HHEAR) laboratory at NYU Langone Medical Center (Dr. Kannan’s laboratory). Fourteen PFAS namely, PFHxS, PFOS, PFOA, PFNA, PFDA, perfluorobutanesulfonic acid (PFBS), perfluoroheptanoic acid (PFHPA), 359 Participants with PFAS Measurements (n=359) Available Fetal Biometry Ultrasound (n=344 participants, 859 scans) No Fetal Biometry Data Available (n=15 participants) Final Sample: 335 participants, 833 scans Missing data on Race/Ethnicity (n=5 participants, 18 scans) Missing data on fetal sex (n=4, 8 scans) 98 perfluroundecanoic acid (PFUNDA), perfluorododecanoic acid (PFDODA), perfluorooctanesulfonamide (PFOSA), n-ethyl perfluorooctane sulfonamido acetic acid (NETFOSAA), n-methyl perfluorooctane sulfonamido acetic acid (NMFOSAA), perfluoro-n- pentanoic acid (PFPEA), and perfluorohexanoic acid (PFHxA) were analyzed. The method for the analysis of 14 PFAS in serum has been previously described [336]. In brief, serum samples (0.25 mL) were aliquoted into 15 mL polypropylene tubes (PP tubes) and spiked with 5 ng of 13 C-labeled internal standard (IS) mixture and 0.7 mL of 1% ammonium formate (w/v) in methanol (MeOH). The mixture was centrifuged for 5 min at 5000 rpm and the supernatant was collected and passed through Hybrid-SPE cartridge (Phospholipid, 30mg, 1cc, Sigma-Aldrich, St. Louis, MO, USA). The cartridges were conditioned with 1 mL of MeOH containing 1% ammonium formate (w/v). The samples were eluted through the cartridge and collected in a PP tube for LC-MS/MS analysis. The target analytes were quantified by isotopic dilution method and a 12-point calibration (at concentrations ranging from 0.02 to 100 nanograms per milliliter (ng/mL)) with the regression coefficient of ≥ 0.999 was used. A pure solvent (MeOH) and a mid-point calibration standard (5 ng/mL) were injected after every 10 samples to check for carry over of target chemicals and instrumental drift in sensitivity. Several procedural blanks were analyzed to monitor for contamination that can arise from reagents and materials used in sample preparation steps. For each batch of 100 samples, five duplicates of procedural blanks and QC spiked samples (water spiked with native standards at 5 ng for all analytes and IS) were processed. Duplicate of Standard Reference Material (SRM1958, NIST, Gaithersburg, MD, USA; IS spiked) containing certified values for PFHxS, PFOS, PFOA, and PFNA were analyzed. Trace levels of NETFOSAA (0.001- 0.047 ng/mL) and NMFOSAA (0.002-0.012 ng/mL) were found in procedural blanks, and the concentrations of these chemicals in samples were subtracted from blank values. Spiked sample 99 and SRM1958 recoveries were in the ranges of 81.5-112% (RSD: ±4.3-8.7%) and 86.7-112% (RSD: ±2.1-6.8%), respectively. The limit of detection (LOD) of target analytes ranged from 0.02 to 0.05 ng/mL. Five of the fourteen PFAS had at least 50% of samples above the LOD and were included in subsequent data analysis approaches. The analytes included were PFOS (100% detected), PFHxS (100% detected), PFNA (70% detected), PFOA (65% detected) and PFDA (57% detected). Fetal Growth Assessment Fetal growth was assessed by transabdominal ultrasound and included five fetal biometry outcomes. Outcomes comprised of two head measurements: 1) head circumference (mm) defined as the length along the skull bone and 2) biparietal diameter (mm) defined as the maximum diameter of a transverse section of the skull. Additionally, we measured femur bone length (mm) and abdominal circumference (mm). These fetal biometry measurements were used to estimate fetal weight (g) using the Hadlock et al. formula [337]. Ultrasound data were obtained from two sources. Licensed, certified sonographers conducted study-measured ultrasound scans at the third trimester study visit (≥ 28 weeks gestation) either in the MADRES study clinic by a single technician (86 scans) or at Keck Hospital by two technicians (61 scans). We also abstracted prenatal electronic medical records (EMR) (686 scans). Scans were conducted throughout pregnancy ranging from 9.6 weeks gestation to 41.6 weeks gestation with the majority in the second and third trimesters. Data were from 10 scans taken in the first trimester (<14 weeks gestation), 400 scans in the second trimester (≥14 weeks to <28 weeks gestation) and 423 scans in the third trimester (≥ 28 weeks gestation). The number of 100 scans per participant ranged from 1-8 with a mean of 2.5± 1.3. All scans were conducted between December 2015 and August 2019. Perceived Stress Scale The 10-item Perceived Stress Scale (PSS) [338] was collected at each trimester study visit via interviewer-administered questionnaire in either Spanish or English. The scale measures the amount of perceived stress experienced by the participant with questions to capture how uncontrollable, unpredictable, and over-loading one’s life was within the last month (30 days). Each question was answered with a Likert scale ranging from never (score of 0) to very often (score of 4). The individual answers were then totaled for an overall PSS score ranging from 0 to 40, with higher numbers indicating more perceived stress. We selected the trimester-specific PSS score that aligned with the blood collection time point with measured PFAS. Covariates Covariates to be included in multivariable models were selected based on current literature. The variables included demographic characteristics, pregnancy related variables, and aspects related to study design. Demographic variables were race/ethnicity, nativity, level of attained education, annual household income, participant’s age at recruitment, pre-pregnancy body mass index (BMI), and any personal reported smoking during pregnancy. All variables were self-reported via interviewer- administered questionnaires in either English or Spanish with the exception of pre-pregnancy BMI in kg/m 2 , which was calculated using self-reported pre-pregnancy weight and study measured height with a stadiometer (Perspectives enterprises model PE-AIM-101). Pregnancy related 101 variables included birth order, which was self-reported via interviewer administered questionnaire, fetal sex, which was abstracted from electronic medical records for the majority of participants (96%) or self-reported from the mother (4%) and fish consumption during pregnancy. Dietary fish intake was captured through an interviewer-administered questionnaire collected at the third trimester visit which accessed the frequency of consuming seafood during pregnancy, including fish sticks, fresh oily fish, other fresh fish, canned tuna, and shellfish. A summary variable was created for all fish intake with the following categories: never (62%), monthly (16%), at least weekly (12%), or unknown (10%). Lastly, variables related to the study design included site of recruitment, ultrasound scan location/technician, gestational age at time of ultrasound scan, and gestational age at time of blood sample for PFAS concentration analysis. Gestational ages at time of blood sample collection and ultrasound scan were calculated by subtracting the number of weeks between the infant’s date of birth and the date of the ultrasound or blood sample from the gestational age in weeks at time of birth. Gestational age at birth was calculated and standardized using a hierarchy of methods [339]. A first trimester (<14 weeks gestation) ultrasound measurement of crown-rump length was deemed ideal if available (60%) and if missing, a second trimester (<28 weeks gestation) ultrasound measurement of fetal biparietal diameter was used (27%). If no measurements from an early ultrasound were available, gestational age at birth was established from a physician’s best clinical estimate from the EMR (13%). Covariates were visualized through a directed acyclic graph (DAG) using DAGitty [340] and minimal sufficient adjustment sets for estimating the total effect of prenatal PFAS exposure on in utero fetal growth consisted of fetal sex, fish consumption, gestational age at blood sample, 102 gestational age at ultrasound, household income, maternal education, nativity, parity, pre- pregnancy BMI and race/ethnicity (Supplemental Figure 1). Statistical Analysis Distributions of participant demographic and pregnancy characteristics were summarized using means and standard deviations for continuous variables (age, pre-pregnancy BMI, gestational age at blood sample, gestational age at ultrasound scan) and frequencies and percentages for categorical variables (income, education, race/ethnicity, fetal sex, parity, fish consumption, nativity). Each fetal growth outcome was visualized through spaghetti plots and smoothed scatterplots across gestational weeks to assess growth trends over time. Linear mixed models were used to evaluate the relationship between PFAS concentrations and each fetal growth outcome. For PFOS and PFHxS, which were detected in 100% of samples, both exposures were modeled continuously with the natural log transformation due to right skewness and to compare results across previous studies. For PFOA, PFNA and PFDA, which had detection percentages ranging from 57%-70%, exposures were dichotomized and modeled categorically as detected (>LOD) or non-detected (≤LOD). Likelihood ratio tests of nested models were used to determine whether a random slope of gestational age at ultrasound scan was needed in addition to the random intercept for each participant. There was strong evidence to include both random intercepts and slopes (p<0.001) and to allow correlation between them (p<0.001). Additionally, due to the non-linearity of fetal growth outcomes over gestation, gestational age at scan also included a quadratic term, which was significant when comparing nested models (p<0.001). We found no evidence for a statistical interaction between PFAS analytes and gestational age at time of ultrasound scan (p>0.05). 103 Final models allowed for a quadratic growth curve in gestational age at time of scan, with participant-level random effects, a main fixed effect of PFAS and adjustment for key covariates. Covariates were modeled as follows: gestational age at scan (weeks, centered at 20 weeks), participant age (years), pre-pregnancy BMI (kg/m 2 ), gestational age at blood sample (weeks), fetal sex (male, female), parity (first born, second or more, missing indicator), race/ethnicity (Hispanic, non-Hispanic Black, non-Hispanic white, non-Hispanic other), annual household income (<$50,000, $50,000-$99,999, ≥$100,000, reported “Don’t Know”), education (high school diploma or less, some or completed college, some graduate training), nativity (US born, non-US born, missing indicator), fish consumption during pregnancy (never, monthly, at least weekly, missing indicator), recruitment site indicator and ultrasound source indicator. Personal smoking during gestation was not included as a covariate due to very few participants reporting any smoking (<2%); however, we evaluated the influence of personal smoking by removing these participants in a sensitivity analysis. To determine if the association of prenatal PFAS exposure on fetal growth trajectories was modified by perceived stress, an interaction term was added to fully adjusted models, and we stratified by higher and lower perceived stress. PSS scores were dichotomized at the median score (PSS Median=13, Range=0-33); 166 participants (totaling 418 scans) were classified as having “lower perceived stress”, and 169 participants (totaling 415 scans) were classified as having “higher perceived stress”. Several sensitivity analyses were conducted to further evaluate the robustness of results. We first removed scans that were taken prior to the blood sample that measured PFAS (46 participants, 212 scans removed) to assess whether PFAS concentrations measured after the ultrasound scans influenced our results. Additionally, we wanted to determine whether maternal 104 pregnancy complications influenced our results. We first additionally adjusted for existing or gestational hypertension/preeclampsia and existing or gestational diabetes that were abstracted from the EMR. Within the sample, 45.3% of participants had either existing or gestational diabetes (33.4%), had either existing or gestational hypertensive disorders (20.3%), or both (8.4%). We then additionally removed participants who had more than five ultrasound scans, which may indicate a higher risk pregnancy (9 participants, 60 scans removed). Lastly, to ensure participants who smoked were not driving our results, participants who reported any smoking during gestation were removed from our models (6 participants, 9 scans removed). Data management was conducted in SAS Version 9.4 and linear mixed models were estimated using the nlme R package [341]. All models met assumptions and were assessed with two-sided hypotheses with 𝛼 = 0.05. Results Participant Characteristics Participants were on average 29±6 years of age at study recruitment, Hispanic (76%), had a high school diploma or less of education (53%), US born (51%) and overweight prior to pregnancy (mean pre-pregnancy BMI: 28±6). The majority of fetuses were male (52%) and were the second or greater child of the participant (62%). Demographic and participant characteristics are shown in Table 1. 105 Table 1. Participant Characteristics (N=335) Characteristic N (%) or Mean (SD) Maternal Age (years) 29.4 (6.0) Race/Ethnicity Hispanic Non-Hispanic Black Non-Hispanic White Non-Hispanic Other 254 (75.8%) 35 (10.5%) 31 (9.2%) 15 (4.5%) Household Income <$50,000 $50,000-$99,999 >$100,000 Reported “Don’t Know” 159 (47.5%) 63 (18.8%) 29 (8.6%) 84 (25.1%) Education Completed high school or less Some college of completed college Some graduate training 177 (52.8%) 129 (38.5%) 29 (8.7%) Country of Birth US Born Non-US Born Unknown 170 (50.7%) 155 (46.3%) 10 (3.0%) Pre-Pregnancy BMI (kg/m 2 ) 28.1 (6.2) Any Personal Smoking During Pregnancy 6 (1.8%) Fetus Male 173 (51.6%) Birth Order First Second or more Unknown 116 (34.6%) 208 (62.1%) 11 (3.3%) Prenatal Serum PFAS Concentrations Five of the fourteen analytes, PFOS, PFHxS, PFNA, PFOA, and PFDA, had at least 50% of samples above the LOD. Of the remaining analytes, the highest percentage of detected samples was 17%. Median concentrations of PFOS, PFHxS, PFNA, PFOA and PFDA were 1.34, 1.10, 0.07, 0.12, 0.04 ng/mL, respectively. The distributions of all measured PFAS concentrations are shown in Table 2. 106 Table 2. Distribution of PFAS (ng/mL) concentrations in maternal blood serum (N=335) Analyte Abbreviation LOD (ng/mL) % Above LOD Min Q 1 Q2 Q 3 Max Perfluorooctanesulfonic acid PFOS 0.02 100% 0.09 0.96 1.34 1.89 10.36 Perfluorohexanesulfonic acid PFHxS 0.02 100% 0.36 0.79 1.10 1.50 4.10 Perfluorononanoic acid PFNA 0.02 70% ND ND 0.07 0.19 1.10 Perfluorooctanoic acid PFOA 0.035 65% ND ND 0.12 0.39 3.43 Perfluorodecanoic acid PFDA 0.035 57% ND ND 0.04 0.09 2.32 Perfluoro-n-pentanoic acid PFPEA 0.05 17% ND ND ND ND 1.19 Ethyl Perfluorooctane sulfonamido acetic acid NETFOSAA 0.02 12% ND ND ND ND 0.12 N-methyl Perfluorooctane sulfonamido acetic acid NMFOSAA 0.02 10% ND ND ND ND 0.38 Perfluroundecanoic acid PFUNDA 0.02 9% ND ND ND ND 0.72 Perfluorohexanoic acid PFHXA 0.05 1% ND ND ND ND 0.11 Perfluorobutanesulfonic acid PFBS 0.02 1% ND ND ND ND 0.08 Perfluorododecanoic acid PFDODA 0.035 1% ND ND ND ND 0.94 Perfluorooctanesulfonamide PFOSA 0.02 0% ND ND ND ND ND Perfluoroheptanoic acid PFHPA 0.05 0% ND ND ND ND ND ND=Not Detected; Q1= Quartile 1 (25 th percentile), Q2= Quartile 2 (50 th percentile), Q3= Quartile 3 (75 th percentile) Prenatal Serum PFAS and Fetal Growth Outcomes We found that participants who had detectible levels of prenatal PFOA exposure had fetuses with -2.5 mm (95% CI -4.2, -0.8) smaller head circumference on average compared to fetuses of participants without detected prenatal PFOA levels for a fixed gestational age at time of ultrasound. Additionally, participants who had detected levels of prenatal PFOA exposure had fetuses with –0.7 mm (95% CI -1.3, -0.2) smaller biparietal diameter compared to fetuses of participants without detected prenatal PFOA levels for a fixed gestational age at time of ultrasound. We did not find significant associations with other PFAS and fetal growth outcomes. We evaluated whether participants experiencing higher compared to lower levels of perceived stress were more vulnerable to the effects of PFOA. In stratified analyses, we found that the association between PFOA and biparietal diameter or head circumference was stronger and only significant in participants experiencing higher levels of stress (Head Circumference: 𝛽 = -3.5 mm, 95% CI -5.8, -1.3; Biparietal Diameter 𝛽 = -0.8 mm, 95% CI -1.6, -0.03) compared to those experiencing lower levels of stress (Figures 2 and 3). A formal test of interaction was not 107 statistically significant (p>0.05). Supplemental Figures 2, 3, and 4 show the remaining non- significant results for prenatal PFAS exposures and other fetal growth outcomes. Lastly, we assessed the robustness of our results of PFOA on fetal head circumference and fetal biparietal diameter through sensitivity analyses and our conclusions remained consistent. Models that 1) additionally adjusted for hypertension/diabetes, 2) removed participants who had more than five ultrasound scans, 3) removed participants with any reported personal gestational smoking, and 4) removed scans that had been conducted before the blood sample was collected are shown in Supplemental Figure 5. Figure 2. Prenatal Serum PFAS Concentrations and Fetal Head Circumference (N=799 scans) All models adjusted for GA at ultrasound scan, GA 2 , fetal sex, parity, education, household income, GA at time of blood sample, maternal race/ethnicity, recruitment site, age at recruitment, pre-pregnancy BMI, source of ultrasound, fish consumption, nativity. Note: PFDA, PFNA and PFOA modeled as >LOD referenced to <=LOD; PFHxS and PFOS modeled continuously and are log transformed. 108 Figure 3. Prenatal Serum PFAS Concentrations and Fetal Biparietal Diameter (N=806 scans) All models adjusted for GA at ultrasound scan, GA 2 , fetal sex, parity, education, household income, GA at time of blood sample, maternal race/ethnicity, recruitment site, age at recruitment, pre-pregnancy BMI, source of ultrasound, fish consumption, nativity. Note: PFDA, PFNA and PFOA modeled as >LOD referenced to <=LOD; PFHxS and PFOS modeled continuously and are log transformed. Discussion This study found that participants with detected levels of PFOA measured during gestation had fetuses with reduced growth trajectories for both head circumference and biparietal diameter when compared to participants who did not have PFOA detected in their blood serum. Additionally, associations were modified by maternal perceived stress during pregnancy with significant inverse associations only observed in mothers reporting higher levels of perceived stress. We also found that many PFAS compounds analyzed within this MADRES sample had limited to no detection and were lower than levels in other pregnancy cohorts with the exception of PFHxS, which was typically higher [328, 329, 342]. Reaching adequate fetal growth milestones across gestation is a critical component of a healthy pregnancy and reduced fetal growth has been shown to predict later life adverse outcomes. 109 Several studies have shown larger in utero fetal head circumference, still in normal ranges, to be associated with positive childhood neurodevelopment reflected through cognitive, language and visual skills, as well as intelligence quotient (IQ) [323, 343, 344]. Research has suggested that environmental exposures in pregnancy could disrupt fetal growth, potentially leading to negative downstream effects [345]. Our results are consistent with other studies that have quantified the association of prenatal PFOA exposure on head circumference at birth [346, 347]. Two studies have explored the association of prenatal PFAS exposure on fetal growth measured in utero and the results have been conflicting [328, 329]. A multi-site US based cohort among higher education and higher income women with low-risk pregnancies assessed 11 PFAS analytes and found prenatal PFAS exposure to be significantly associated with femur length, although depending on the analyte, both positive (NMeFOSAA, PFDA, PFHpA, PFHxS, PFNA, PFOA) and negative associations (PFOSA, PFDS) were observed [329]. No significant associations were observed with biparietal diameter or fetal head circumference [329]. A study in Spain among higher education and higher income women found no significant main effects between PFAS exposure and fetal growth outcomes, but showed evidence for effect modification by maternal smoking status with smoking mothers having significantly decreased fetal femur length and estimated fetal weight with PFOA and PFNA exposure [328]. No significant associations were observed with fetal biparietal diameter, and head circumference was not included within the study [328]. Although there is evidence that psychosocial stress and chemical exposures during pregnancy may produce joint effects on fetal growth [334], no studies to date have assessed whether maternal perceived stress modifies the influence of prenatal PFAS exposure on fetal growth measured in utero. One study used a mixture approach of prenatal PFAS exposure and 110 maternal stressors (perceived stress and depression) through quantile g-computation and observed an inverse association with the PFAS and stress mixture and birthweight z-scores, although it did not reach statistical significance [348]. Two studies have shown that prenatal PFAS exposure influences corticotrophin-releasing hormone and cortisone levels in pregnancy [349, 350], suggesting that stress may be an important effect modifier for PFAS exposure effects. Future studies should further investigate how maternal stress and PFAS interact to impact fetal growth, particularly in populations experiencing greater health disparities who also tend to have higher exposures to environmental contaminants and may face a unique constellation of stressors [331, 351]. The current study has several strengths. First, the study population is comprised of predominately low-income Hispanic women who traditionally have been excluded from previous research and are underrepresented in the literature. Additionally, several PFAS were assessed, including PFNA, PFDA, and PFHxS, which have not been as comprehensively studied as PFOA and PFOS. This study also examined the relationship of these compounds on longitudinal fetal growth trajectories measured across pregnancy, rather than on birth outcomes, which are commonly used as proxies for fetal growth. Lastly, this is one of the first studies to explore how perceived stress experienced in pregnancy modifies the association with PFAS on fetal growth and suggests that women with higher stress and higher PFOA levels face larger adverse impacts on fetal development. Limitations also exist within this study. First, PFAS concentrations were measured only once during pregnancy, which limited our ability to assess how these levels changed across gestation and samples were not limited to a single trimester for blood serum collection. However, given their long half-lives, this study assumed that PFAS levels were relatively stable across 111 pregnancy. A previous study with PFOA and PFOS levels measured from blood serum in both the first and second trimester found a high degree of correlation between the two time points [352]. An additional study with blood samples taken at all three trimesters of pregnancy found PFOA, PFOS, PFNA, and PFDA to slightly decrease over pregnancy while PFHxS remained stable [353]. Additionally, we were limited in our modeling approaches as several analytes measured in the panel were not detected or had very low detection. Nine of the fourteen measured analytes were excluded completely and three of the five included ranged between 57-70% detected. For these reasons, mixture methods approaches were not conducted and PFOA, PFNA, and PFDA were only assessed categorically. Lastly, although several sensitivity analyses were performed to assess the robustness of our results and we included covariates determined to be associated with both PFAS exposure and fetal growth in multivariable models, we acknowledge that residual confounding remains possible. Conclusion Maternal serum PFOA concentrations were associated with lower fetal head circumference and biparietal diameter growth trajectories across pregnancy and these associations were stronger among participants reporting higher levels of perceived stress. 112 Supplemental Material Supplemental Figure 1. Directed Acyclic Graph (DAG) of PFAS Exposure and Fetal Growth Supplemental Figure 2. Prenatal Serum PFAS Concentrations and Fetal Abdominal Circumference in mm (N=800 scans) All models adjusted for GA at ultrasound scan, GA 2 , fetal sex, parity, education, household income, GA at time of blood sample, maternal race/ethnicity, recruitment site, age at recruitment, pre-pregnancy BMI, source of ultrasound, fish consumption, nativity. Note: PFDA, PFNA and PFOA modeled as >LOD referenced to <=LOD; PFHxS and PFOS modeled continuously and are log transformed. 113 Supplemental Figure 3. Prenatal Serum PFAS Concentrations and Fetal Femur Length in mm (N=800 scans) All models adjusted for GA at ultrasound scan, GA 2 , fetal sex, parity, education, household income, GA at time of blood sample, maternal race/ethnicity, recruitment site, age at recruitment, pre-pregnancy BMI, source of ultrasound, fish consumption, nativity. Note: PFDA, PFNA and PFOA modeled as >LOD referenced to <=LOD; PFHxS and PFOS modeled continuously and are log transformed. Supplemental Figure 4. Prenatal Serum PFAS Concentrations and Estimated Fetal Weight in g (N=804 scans) All models adjusted for GA at ultrasound scan, GA 2 , fetal sex, parity, education, household income, GA at time of blood sample, maternal race/ethnicity, recruitment site, age at recruitment, pre-pregnancy BMI, source of ultrasound, fish consumption, nativity. Note: PFDA, PFNA and PFOA modeled as >LOD referenced to <=LOD; PFHxS and PFOS modeled continuously and are log transformed. 114 Supplemental Figure 5. Sensitivity Analyses for Prenatal PFOA Exposure and Biparietal Diameter (BPD) and Head Circumference (HC) All models adjusted for GA at ultrasound scan, GA 2 , fetal sex, parity, education, household income, GA at time of blood sample, maternal race/ethnicity, recruitment site, age at recruitment, pre-pregnancy BMI, source of ultrasound, fish consumption, nativity. Note: PFOA modeled as >LOD referenced to <=LOD. 115 CHAPTER SIX Study 3: Defining Windows of Susceptibility to Ambient Air Pollution Exposure with Greatest Effects on Fetal Growth: The MADRES Cohort Abstract Background It is well documented that persons of color experience disproportionately higher exposure to environmental contaminants, including air pollution, and have poorer pregnancy outcomes. This study assessed the critical windows of exposure to prenatal ambient air pollution on in utero fetal growth among structurally marginalized populations in urban Los Angeles, California. Methods Participants (N=281) from the larger ongoing Maternal And Developmental Risks from Environmental and Social Stressors (MADRES) pregnancy cohort study who had an ultrasound in the third trimester were included in this analysis. Fetal growth outcomes were measured on average at 32±2 weeks of gestation by a certified sonographer and included estimated fetal weight, head circumference, abdominal circumference, biparietal diameter and femur length. Daily ambient air pollutant concentrations were estimated for four pollutants (particulate matter less than 2.5 µm (PM2.5) and less than 10 µm (PM10) in aerodynamic diameter, nitrogen dioxide (NO2), and 8-hour maximum ozone (O3)) at participant residences using inverse-distance squared spatial interpolation from ambient monitoring data. Weekly gestational averages were calculated from daily concentrations from 12 weeks prior to conception to 32 weeks of gestation (44 total weeks), and their associations with growth outcomes were estimated using distributed lag models (DLMs) adjusted for maternal age, race, ethnicity, education level, household income, parity, pre- 116 pregnancy body mass index (BMI), fetal sex, gestational age at ultrasound scan, ultrasound technician, lag-specific averaged temperature, and season of ultrasound. Results Participants were on average 29±6 years old, overweight prior to pregnancy (BMI=29±7 kg/m 2 ), predominately Hispanic (82%) and the majority had a high school diploma or less in education (58%). We identified a significant sensitive window of PM2.5 exposure (per IQR increase of 6 μg/m 3 ) between gestational weeks 4-16 for lower estimated fetal weight (𝛽 averaged4-16 =-8.7 g; 95% CI -16.7, -0.8). Exposure to PM2.5 during gestational weeks 1–23 was also significantly associated with smaller fetal abdominal circumference (𝛽 averaged1-23 =-0.6 mm; 95% CI -1.1, -0.2). Additionally, prenatal exposure to PM10 (per IQR increase of 13 μg/m 3 ) between weeks 6–15 of pregnancy was significantly associated with smaller fetal abdominal circumference (𝛽 averaged6-15 = -0.4 mm; 95% CI -0.8, -0.1). No other significant windows of exposure from preconception to late pregnancy were observed. Conclusions These results suggest that exposure to particulate matter in early to mid-pregnancy, but not preconception or late pregnancy, may have critical implications on fetal growth. Introduction Ambient air pollution consists of a diverse mixture of gases and particles and has the ability to potentially affect every organ within the human body [354]. Air pollutants are known contributors to cardiovascular disease and premature mortality with an estimated 135 million 117 Americans currently affected by unhealthy levels of ozone and particulate matter (PM) [355-358]. Additionally, air pollution exposure has been shown to differ dramatically based on race and ethnicity within the United States. A recent study found that exposure to PM with an aerodynamic diameter of 2.5 micrometers (μm) or less (PM2.5), is greater for people of color even after accounting for state, geographic area (urban vs. rural), and income level [359]. Persons from racial and ethnic minority groups experience disproportionate exposures to toxic air emissions due to residential proximity to industrial practices [360, 361] and are more likely to live near traffic sources compared to non-Hispanic whites [362]. In addition to experiencing higher exposures to air pollution, communities of color also tend to have higher rates of adverse birth outcomes compared to non-Hispanic whites [363]. A myriad of complex factors contribute to these health disparities including racial discrimination, less access to prenatal care, and lower socio-economic status [364, 365]. Another driving factor may be higher levels of exposure to air pollution during pregnancy. Substantial evidence documents that exposure to ambient air pollution during pregnancy influences negative birth outcomes including low birthweight (<2,500g) and preterm birth (<37 weeks) [366-373] due primarily to biological mechanisms impacting the placenta [374-377]. Adverse negative birth outcomes, and decreased fetal growth, influences health across the life course [378-380]. Fetal growth in utero is commonly assessed by sonographic measurements of the fetus’s head circumference, biparietal diameter, abdominal circumference, femur length as well as estimated fetal weight. Although there is extensive literature on the influence of air pollution on birth outcomes and infant birthweight, there are far fewer studies that have assessed fetal growth and development and the majority of these studies are conducted in more affluent populations. A study in Spain found that nitrogen dioxide (NO2) exposure during pregnancy was associated with 118 decreased biparietal diameter, abdominal circumference and estimated fetal weight [381-383]. NO2 has also been associated with significant inverse effects on fetal growth in cohorts in Los Angeles, California [384], China [385], and the Netherlands [386]. Additionally, studies have shown PM with an aerodynamic diameter of 10 μm or less (PM10) [384, 386-390] and PM2.5 [387, 391-393] to have adverse effects on fetal growth. Studies investigating ground level ozone (O 3) are limited [384, 388] with only one finding significant inverse effects on fetal abdominal circumference [288]. While the literature suggests air pollution adversely impacts in utero fetal growth, direct comparison of results across studies is challenging due to differences in growth outcomes measured, timing of measurements, and differences in exposure assessment and modeling approaches. None of the existing studies addressed finer time resolution of exposure (i.e. weekly) that may be critical for identifying subtle, but important, in utero growth effects. Averaging exposure across pregnancy lacks the ability to identify critical windows of exposure or capture the true variability of the exposures across gestation. Exposure time windows have also generally been limited to prenatal time points, with few studies exploring whether preconception exposures influence fetal growth. Preconception air pollution has been shown to influence other pregnancy outcomes including gestational diabetes [394-396]. To address these gaps, this study assessed the influence of preconception and in utero exposure to four criteria air pollutants including PM2.5, PM10, NO2, and (8-hour maximum) O3 on third trimester fetal growth within a structurally marginalized population of predominately Hispanic participants. We hypothesized that there are important windows of susceptibility to ambient air pollution exposure that are associated with reduced fetal weight, abdominal circumference, femur length, biparietal diameter, and head circumference. We explored this using 119 distributed lag models (DLMs) of exposures spanning from 12 weeks preconception to 32 weeks of gestation. As a secondary aim, we also assessed prenatal air pollution on in utero fetal growth using standard pregnancy-average pollution exposure effects within this population. Methods Sample Participants are drawn from the larger ongoing Maternal And Developmental Risks from Environmental and Social Stressors (MADRES) pregnancy cohort study. An overview of the study design, protocol, and demographics of the cohort have been previously described [397]. In brief, participants were enrolled into the study during pregnancy from four prenatal clinic sites. The sites include two community health clinics, one county hospital prenatal clinic, and one private obstetrics and gynecology practice. Eligibility criteria for cohort entry included: (1) <30 weeks pregnant, (2) ≥18 years of age, and (3) fluent in English or Spanish. Exclusion criteria for the study included: (1) multiple gestation; (2) having a physical, mental, or cognitive disability that would inhibit participation or the ability to provide consent; (3) current incarceration; or (4) HIV positive status. Informed consent and HIPAA authorization for medical record abstraction were obtained from each participant at time of study entry and the University of Southern California’s Institutional Review Board (IRB) approved all study aspects. This study included participants who had a study-measured ultrasound conducted by a licensed sonographer during the third trimester visit (≥28 weeks gestation). Of the 285 participants with a fetal ultrasound in the third trimester, one participant was missing data on prenatal ambient air pollution, and an additional three participants were missing information on race/ethnicity leaving a final sample size of 281 participants. The consort diagram is presented in Figure 1. 120 Participants had similar sociodemographic characteristics to the overall MADRES cohort (Supplemental Table 1). Figure 1. Consort Diagram of Included Participants Prenatal Ambient Air Pollution Measurements Daily residential histories were assembled and geocoded for each participant using residential address and occupancy dates collected with residential history forms and prospective address confirmation data collected at every contact point. These capture all residential mobility or moves and form the basis of all geospatial exposure assignments. Daily ambient air pollutant concentrations of PM2.5, PM10, O3 (8-hour maximum) and NO2, were then assigned using inverse- distance-squared weighted spatial interpolation from the United States Environmental Protection Agency (EPA) Air Quality System monitors. For PM2.5, PM10, and NO2, 24-hour daily averages were used, while the 8-hour daily maximum was used for O3. Additionally, daily temperature in degrees Celsius was calculated as the average between the minimum and maximum temperature, which were obtained from a high-resolution (4 km x 4 km) gridded meteorological dataset [398]. Weekly averages were calculated from daily 285 MADRES Participants with a 3 rd Trimester Ultrasound 284 with Ambient Air Pollution Data 281 with Key Covariate Data N=281 1 participant missing prenatal air pollution data 3 participants missing data for race/ethnicity 121 concentrations of pollutants and temperature from 12 weeks prior to conception until 32 weeks gestation (44 total weeks). The weekly averages were computed with generally complete data (<1% of daily concentrations were missing). Three participants had missing data on pollutants for certain weeks during pregnancy (4-22 total weeks) due to inadequate quality of address geocode match or living abroad during a period of gestation. We chose to examine preconception windows up to 12 weeks prior to conception due to considering this as the biologically relevant window of exposure and previous literature using this time period for risk of gestational diabetes [389-391]. Exposures through 32 weeks gestation were chosen to correspond to the mean gestational age at ultrasound measurement across participants. Overall pregnancy average exposures were calculated for each participant starting from date of conception until the date of the ultrasound scan (consisting of a mean of 224±12 days). Fetal Growth Measurements Five fetal growth outcomes were measured via transabdominal ultrasound. Outcomes included head circumference in millimeters (mm) defined as the length along the skull bone, biparietal diameter (mm) defined as the maximum diameter of a transverse section the fetal skull from the proximal parietal bone to the inner edge of the distal parietal bone, femur length (mm) defined as the length of the thigh bone, abdominal circumference (mm) which is measured at the widest part across the fetal liver and estimated fetal weight in grams (g) derived from the formula from Hadlock et al., which takes into account the above measurements [399-402]. This analysis used fetal biometry measurements from a single study-measured ultrasound scan (Mean: 32±2; Range: 28-36 weeks gestation) conducted by certified, licensed sonographers at the third trimester study visit. A total of 65 scans were conducted from August 2016 to August 2018 by two sonographers at Keck Hospital of USC with a Toshiba Aplio 500 machine. The 122 sonographers conducted roughly equal numbers of scans (“sonographer A” 30 scans, “sonographer B” 35 scans). The remaining 216 scans were conducted by a single sonographer (“sonographer C”) at the MADRES clinic from August 2018 to March 2020 using a Philips CX-50 machine, with a convex 1-5MHz transducer. Covariates Covariates to include in multivariate models were identified a priori based on a review of the literature and visualized through Directed Acyclic Graphs (DAGs) [403]. Maternal and demographic covariates were included such as maternal age at time of study recruitment, maternal race and ethnicity, maternal education level, annual household income, parity, pre-pregnancy body mass index (BMI), fetal sex, gestational age at time of ultrasound scan, average temperature, season of ultrasound, recruitment site, and ultrasound technician. We additionally considered personal smoking during pregnancy, but due to very few participants reporting any personal smoking during pregnancy (<2.5%) it was not included as a covariate and was instead evaluated in a sensitivity analysis by removing smoking participants. Age, race/ethnicity, education, household income, birth order of the child, and smoking were self-reported via interviewer-administered questionnaires in English or Spanish. Pre- pregnancy BMI was computed using self-reported pre-pregnancy weight and study measured standing height using a stadiometer (Perspectives enterprises model PE-AIM-101). Sex of the fetus was abstracted from the EMR (98%) or came from the birth information form/proxy report from the mother (2%) and gestational age at time of ultrasound scan in weeks (including partial weeks by number of additional days) was calculated by subtracting the difference in number of weeks between the infant’s date of birth and the date of the ultrasound from the gestational age in weeks at time of birth. Gestational age at birth was calculated and standardized using a hierarchy of 123 methods [404]. A first trimester (<14 weeks gestation) ultrasound measurement of crown-rump length was considered highest quality and was used if available (59%). If missing, a second trimester (<28 weeks gestation) ultrasound measurement of fetal biparietal diameter was used (27%). If measurements from an early ultrasound were missing, gestational age at birth was established from a physician’s best clinical estimate from the EMR (14%). Statistical Analysis Distributions of participant demographic and health characteristics were summarized using means and standard deviations for continuous variables and frequencies and percentages for categorical variables. Due to right skewed distributions, medians and interquartile ranges (IQR) were computed for each ambient air pollutant and we used Spearman correlations to assess the relationships between the air pollutants. Spaghetti plots were generated to visualize the change in pollutant exposure across the weeks of gestation for each participant. Means and standard deviations were calculated for each fetal growth outcome and Pearson correlations were computed to assess the relationships across the fetal biometry measurements due to meeting parametric assumptions. We fit DLMs [405] to estimate the time-varying associations between air pollutants at each week, from 12 weeks preconception to 32 weeks of gestation (44 total weeks), to determine critical windows of exposure. This approach allowed the effects of air pollution and temperature to be distributed across time with average weekly levels created using a “cross-basis” function which constrained correlation across weeks [405]. In the DLMs, natural cubic splines with 2-8 degrees of freedom were tested while adjusting for covariates. Covariates were modeled as follows: age in years, pre-pregnancy BMI in kg/m 3 , gestational age at time of scan in weeks, week-specific average temperature in degrees Celsius, race/ethnicity (Hispanic, non-Hispanic Black, non- 124 Hispanic White, non-Hispanic Other), education (high school diploma or less, some or completed college), household income (<$50,000, ≥$50,000, or reported “Don’t Know”), birth order (first, second or more, missing data indicator), fetal sex (male or female), season at time of ultrasound scan (winter, fall, spring, or summer), technician, and recruitment site. The model with the lowest Akaike information criteria (AIC) was chosen and knots were placed at weeks 2 and 20 of gestation. The regression slope for weekly air pollution exposure was scaled to an IQR increase for the pollutant. A sensitive window was defined as week(s) when the 95% confidence intervals (CI) did not include zero. Effect modification by fetal sex was evaluated through stratified models. Sensitivity analyses were conducted to further evaluate robustness of results by first removing participants who reported any smoking during pregnancy (<2.5%) and then separately by adjusting for high-risk pregnancies captured by gestational or chronic diabetes and hypertension reported on EMR (49%). DLMs were fit using the dnlm package in R [405]. Although this package allows for non-linear modeling approaches, we only considered linear air pollution effects at a given lag, which was confirmed appropriate by spline tests using generalized additive models (GAM) with overall prenatal air pollution and each growth outcome (p>0.05). To compare to more traditionally used linear regression results, a secondary analysis was conducted using the overall pregnancy average concentration of each pollutant from date of conception until date of the ultrasound scan as the exposure with each of the fetal biometry outcomes. Linear regression models were adjusted for the same covariates. Beta estimates for the exposure were standardized to the pollutant’s IQR in order to aid with interpretation. Additionally, multipollutant models were then conducted to assess robustness of single pollutant results. Influential points were assessed through Jackknife residuals and Cook’s D and all models met the assumptions of linear regression. Data cleaning, management, and linear regression models were 125 conducted in SAS Version 9.4 and DLM analyses were conducted with 4.04. Version of R. All analyses were conducted with two-sided hypotheses and an alpha level of 0.05. Results Participant Characteristics Participants were on average 29±6 years of age, primarily overweight prior to pregnancy (mean BMI=29±7), predominately Hispanic (82%), nearly half had family incomes of less than $50,000 per year (44%), and a majority of participants had a high school diploma or less in education (58%). There were slightly more female fetuses (52%), and the majority of participants were pregnant with at least their second child (69%). Participant characteristics are shown in Table 1. Table 1. Participant Characteristics (N=281) Characteristic Mean (SD) or N (%) Participant Race/Ethnicity Hispanic Non-Hispanic Black Non-Hispanic White Non-Hispanic Other 230 (81.9%) 36 (12.8%) 8 (2.8%) 7 (2.5%) Income < $50,000 ≥$50,000 Reported “Don’t Know” 123 (43.8%) 52 (18.5%) 106 (37.7%) Education HS Diploma or Less Some or Completed College 164 (58.4%) 117 (41.6%) Age (years) 28.5 (6.1) Pre-Pregnancy BMI 29.2 (6.7) Any Prenatal Smoking 7 (2.5%) Fetus Sex Female 146 (52.0%) Birth Order First Second or more Unknown 80 (28.5%) 195 (69.4%) 6 (2.1%) 126 Ambient Air Pollution Median and IQR of averaged PM2.5, PM10, NO2, and O3 concentrations from date of conception to date of ultrasound and median and IQR of averaged weekly (12 weeks preconception to 32 weeks of gestation) concentrations are shown in Table 2. Overall pregnancy average 8-hr maximum O3 concentrations were inversely correlated with 24-hr NO2 (𝜌 = -0.61, p<0.001) and 24-hr PM2.5 (𝜌 = -0.16, p=0.007), while positively correlated with 24-hr PM10 (𝜌 = 0.22, p=0.0002). NO2 was positively correlated with both PM2.5 (𝜌 =0.46, p<0.0001) and PM10 (𝜌 = 0.20, p=0.001). PM2.5 and PM10 were positively correlated with one another (𝜌 =0.64, p<0.0001). Plots of weekly pollutant correlations (Week 1 to Week 32) are shown in Supplemental Figure 1. Table 2. Distributions of Pregnancy Average and Weekly Average Ambient Air Pollutants Pollutant Pregnancy Average Median (IQR) Weekly Average Median (IQR) PM2.5 (𝜇 g/m 3 ) 11.7 (2.2) 11.7 (1.9) PM10 (𝜇 g/m 3 ) 27.9 (7.1) 29.5 (6.5) NO2 (ppb) 15.5 (5.6) 15.3 (5.6) O3 (ppb) 42.3 (5.7) 42.9 (5.6) *Note: PM2.5, PM10, and NO2 are from 24-hour estimates, O3 represents the 8-hour maximum Pregnancy average: from date of conception to date of ultrasound scan Weekly average: from 12 weeks preconception to 32 weeks of gestation Fetal Growth Outcomes Fetal growth was measured via ultrasound at a mean of 32±2 weeks gestation and the majority of scans were conducted in Fall or Winter (32% Fall, 31% Winter, 19% Spring, 18% Summer). Fetal biometry outcomes were normally distributed (Shapiro Wilk p values >0.05) with the exception of estimated fetal weight, which was slightly left skewed. Means and standard deviations are shown in Table 3. All outcomes were significantly positively correlated with one another (Pearson’s 𝜌 = 0.75 to 0.97, p<0.0001). Fetal growth outcomes had weak to moderate positive significant correlations with infant birth weight. Correlations are shown in Supplemental Figure 2. 127 Table 3. Descriptive Statistics of Fetal Growth Outcomes Fetal Growth Outcome N Mean (SD) Estimated Fetal Weight (g) 281 1957.0 (376.9) Head Circumference (mm) 281 297.7 (14.1) Abdominal Circumference (mm) 281 282.7 (19.9) Biparietal Diameter (mm) 281 80.6 (4.3) Femur Length (mm) 279 61.3 (3.4) Prenatal Air Pollution Exposure and Fetal Growth We investigated weekly lagged exposure from 12 weeks preconception to 32 weeks of gestation. After adjustment for covariates, we found significant windows of exposure in which PM2.5 and PM10 had inverse associations with fetal abdominal circumference and PM2.5 with estimated fetal weight (Figure 2 and 3). We observed a significant sensitive window of PM2.5 exposure (per IQR increase of 6 μg/m 3 ) between weeks 4-16 for lower estimated fetal weight (𝛽 averaged4-16 = -8.7 g; 95% CI -16.7, -0.8). The strongest effect size was at week 8 of pregnancy (𝛽 = -9.3 g; 95% CI -17.3, -1.2). Exposure to PM2.5 during gestational weeks 1–23 was also significantly associated with smaller fetal abdominal circumference (𝛽 averaged1-23 = -0.6 mm; 95% CI -1.1, -0.2) with the strongest effect size at week 11 of pregnancy (𝛽 = -0.7 mm; 95% CI -1.2, - 0.3). Similar associations for PM10 were observed with abdominal circumference (Figure 4). Prenatal exposure to PM10 (per IQR increase of 13 μg/m 3 ) between weeks 6–15 of pregnancy was significantly associated with smaller third trimester fetal abdominal circumference (𝛽 averaged6-15 = - 0.4 mm; 95% CI -0.8, -0.1) with the strongest effect size at week 9 of pregnancy (𝛽 = -0.5 mm; 95% CI -0.9, -0.1). No other ambient air pollutants or fetal growth outcomes showed significant weeks of exposure through DLMs. In stratified models, we found different patterns of associations for PM10 exposure and abdominal circumference by fetal sex. Critical windows of exposure were identified among female fetuses (N=146) between weeks 12-14, while none were observed among male fetuses (N=135) 128 (Supplemental Figure 3). Among female fetuses, a 1 IQR increase in PM 10 was associated with 𝛽 averaged12-14= -0.5 mm (95% CI -1.1, -0.1) smaller fetal abdominal circumference on average. No other significant windows were identified in stratified models. Results were consistent when we removed participants with any reported smoking during pregnancy (N=7) (Supplemental Figure 4) and when we additionally adjusted for maternal health complications (chronic and gestational diabetes/hypertension) (Supplemental Figure 5). Looking at pregnancy-wide exposures and fetal growth using fully adjusted linear regression models, our findings were consistent that particulate matter impacts fetal growth (Table 4). Per one IQR (IQR=2.2 𝜇 g/m 3 ) increase in prenatal average PM2.5 exposure, fetal abdominal circumference decreased by 3.8 mm (95% CI -7.1, -0.4) on average. Additionally, estimated fetal weight decreased by -57.5 g (95% CI -114.5, -0.4) on average. Results remained statistically significant in multipollutant models that additionally adjusted for O3 (PM2.5 and estimated fetal weight 𝛽 = -64.7 g; 95% CI -127.7, -1.6 and PM2.5 and abdominal circumference 𝛽 = -4.2 mm; 95% CI -7.9, -0.5). There were no other significant associations between prenatal averaged ambient air pollutants and fetal growth, although PM10 exposure and fetal abdominal circumference suggested an inverse trend (𝛽 = -1.8 mm per IQR; 95% CI -4.9, 1.3). 129 Figure 2. DLM Results Relating Weekly PM 2.5 Exposure to Estimated Fetal Weight per IQR Note: red estimate p<0.05; IQR=6 μg/m 3 Adjusted for maternal age at time of study recruitment, maternal race and ethnicity, maternal education level, household income, parity, pre-pregnancy body mass index (BMI), sex of the fetus, gestational age at time of ultrasound scan, lag specific average temperature, season of ultrasound, ultrasound technician, and recruitment site. Figure 3. DLM Results Relating Weekly PM 2.5 Exposure to Fetal Abdominal Circumference per IQR Notes: red estimate p<0.05; IQR=6 μg/m 3 Adjusted for maternal age at time of study recruitment, maternal race and ethnicity, maternal education level, household income, parity, pre-pregnancy body mass index (BMI), sex of the fetus, gestational age at time of ultrasound scan, lag specific average temperature, season of ultrasound, ultrasound technician, and recruitment site. 130 Figure 4. DLM Results Relating Weekly PM 10 Exposure to Fetal Abdominal Circumference (mm) per IQR Notes: red estimate p<0.05; IQR=13 μg/m 3 Adjusted for maternal age at time of study recruitment, maternal race and ethnicity, maternal education level, household income, parity, pre-pregnancy body mass index (BMI), sex of the fetus, gestational age at time of ultrasound scan, lag specific average temperature, season of ultrasound, ultrasound technician, and recruitment site. Table 4. Linear Regression Results for Prenatal Average Air Pollution Exposure and Fetal Growth EFW (g) 1 HC (mm) 1 AC (mm) 1 BPD (mm) 1 FL (mm) 1 β (95% CI) β (95% CI) β (95% CI) β (95% CI) β (95% CI) NO 2 (ppb) -35.6 (-98.7, 27.5) -1.2 (-4.0, 1.6) -2.6 (-6.3, 1.2) 0.001 (-0.9, 0.9) 0.1 (-0.6, 0.9) O 3 (ppb) 13.6 (-38.1, 65.2) -0.7 (-3.0, 1.6) 0.8 (-2.3, 3.8) 0.2 (-0.5, 1.0) 0.2 (-0.4, 0.8) PM 2.5 (𝜇 g/m 3 ) -57.5 (-114.5, -0.4)* -1.0 (-3.5, 1.6) -3.8 (-7.1, -0.4)* -0.4 (-1.2, 0.4) -0.3 (-1.1, 0.3) PM 10 (𝜇 g/m 3 ) -22.4 (-75.5, 30.7) 0.3 (-2.0, 2.7) -1.8 (-4.9, 1.3) -0.2 (-0.9, 0.6) 0.3 (-0.5, 0.7) Notes: Beta estimates are standardized to the IQR of the pregnancy average ambient air pollutant NO2 =5.6 ppb; O3=5.7 ppb; PM2.5=2.2 𝜇 g/m 3 PM10= 7.1𝜇 g/m 3 ; *p<0.05; EFW=Estimated Fetal Weight, HC=Head Circumference, AC=Abdominal Circumference, BPD=Biparietal Diameter, FL=Femur Length 1 Adjusted for maternal age at time of study recruitment, maternal race and ethnicity, maternal education level, household income, parity, pre-pregnancy body mass index (BMI), sex of the fetus, gestational age at time of ultrasound scan, average temperature, season of ultrasound, ultrasound technician, and recruitment site. Discussion In this study of primarily low-income Hispanic participants residing in urban Los Angeles, we identified critical windows of prenatal exposure to PM2.5 that were significantly associated with lower fetal growth. Exposure to PM2.5 during weeks 4 to 16 of pregnancy was associated with 131 lower estimated fetal weight and exposure from weeks 1 to 23 of pregnancy was associated with smaller fetal abdominal circumference. The strongest effect sizes were observed at week 8 and 11, respectively. We additionally found that prenatal PM10 exposure from weeks 7 to 15 of pregnancy was associated with smaller fetal abdominal circumference, with the strongest effect size at week 9. Our results are consistent with four studies that have assessed the impacts of prenatal PM2.5 exposure across averaged time windows on in utero fetal growth [387, 391-393]. Leung et al. assessed spatiotemporal modeled PM2.5 exposure and fetal growth among predominately non- Hispanic white participants in Eastern Massachusetts [392]. They found average PM2.5 from conception through 16 weeks of gestation was associated with reduced abdominal circumference measured at both <24 weeks and ≥24 weeks. Lin et al. assessed average ambient PM2.5 exposure from date of last menstrual period to date of ultrasound on estimated fetal weight in a study in Beijing, China, and found reduced fetal weight with increasing PM 2.5 exposure [393]. Cao et al. found average prenatal ambient PM2.5 from conception to one week prior to ultrasound was associated with reduced fetal abdominal circumference and estimated fetal weight in participants in Shanghai, China [391]. Clemens et al. found non-significant inverse associations between annual average PM2.5 concentrations with abdominal circumference in a study in Scotland [387]. Consistent with our results, two previous studies have found averaged prenatal PM 10 exposure to have significant inverse effects on fetal abdominal circumference [388, 389]. Of the studies that have assessed PM10 on fetal abdominal circumference [384, 387-389], none stratified results by fetal sex except Clemens et al. in sensitivity analyses, and results were not shown [387]. Previous research has suggested female fetuses are more susceptible to exposures late in the first trimester and during the second trimester, possibly indicated by more female fetus pregnancy losses 132 compared to male fetuses during this time period [406]. This would overlap with the significant weeks identified in our study. Normal fetal growth across gestation is a critical component of a healthy pregnancy. Intrauterine growth restriction and slow in utero growth can influence the long-term health of the child including increased risk for type 2 diabetes mellitus, coronary heart disease, and hypertension [378-380]. Markers of abnormal gestational growth have been associated with later life outcomes. For example, reduced fetal head circumference has been associated with reduced childhood intelligence quotient (IQ) [407] and smaller fetal abdominal circumference with childhood allergy outcomes [408] and BMI [409]. Suggested biological mechanisms for how prenatal air pollution influences gestation include impacts on the placenta, which is a susceptible target of environmental insults, and disruption of placental function can lead to altered fetal growth [374]. A recent review indicated that prenatal air pollution exposure is associated with both nitrosative stress and epigenetic changes in the placenta [377]. Specifically, PM has been associated with oxidative stress leading to placental inflammation and eventually impaired transplacental oxygen and nutrient delivery to the fetus by triggered hemodynamic responses, ultimately impacting fetal growth [410]. This current study has several notable strengths. MADRES is a well-characterized longitudinal cohort study. The cohort represents a structurally marginalized population who have traditionally been excluded from research and have higher exposure to environmental contaminants due to structural inequities based on race and ethnicity [360, 361]. Within the United States, pregnant persons of color, specifically Black and Hispanic populations, have higher rates of preterm birth and low birthweight when compared to their non-Hispanic white counterparts [363]. Communities of color are also disproportionally burdened by environmental health hazards 133 including air pollution [359, 411]. Previous research has suggested that policies that focus on reducing air pollution exposure have the potential to decrease the Black-white disparities of preterm birth [412]. Another strength of the study was the detailed residential histories, which allowed for finely time-resolved estimates of exposure and accounted for housing relocations during the preconception and gestational time periods. We used these daily estimates to not only create overall averages, which is currently the primary method of exposure modeling within these associations in the literature, but also to create weekly averages to include in DLMs. A key strength of the DLM framework is that it utilizes the data from all weeks concurrently, and assumes that the association varies smoothly as a function of time, while also adjusting for the exposure levels at the different weeks included [413]. This allowed us to identify critical windows of exposure while also not increasing the likelihood for a Type 1 error. Including preconception air pollution exposure in the modeling framework is an additional strength as it is a currently understudied potential critical period of susceptibility for fetal growth effects. Although our study examined exposure to air pollution three months prior to conception, no critical windows of exposure on fetal growth were identified during this preconception period. Emerging literature has shown that preconception air pollution exposure influences pregnancy outcomes, notably gestational diabetes mellitus [394-396, 414]. The influence of preconception exposure to air pollution on fetal growth may be smaller than that of exposure during pregnancy, thus indicating more statistical power may be needed through larger sample sizes. There are currently no other studies that have examined the effects of preconception air pollution on in utero fetal growth, and future studies should explore this further when investigating critical windows of exposure. 134 As in all observational studies, the current study also has limitations to address. Due to the urban environment of Los Angeles, our results may not be generalizable to other regions with a different mixture of air pollution components or those with differing levels. Additionally, although medical record data were available for chronic and gestational health outcomes related to diabetes and hypertension, measurement error in confounders is possible. We also were unable to assess the potential modifying effects of prenatal smoking, due to very few participants reporting any smoking during pregnancy. However, this likely eliminated any confounding effect from personal smoking during gestation, assuming accurate reporting. Our study also only included participants that had ultrasounds taken in the third trimester, which would inherently exclude pregnancies that had ended in early miscarriages. This may also be why we did not see critical windows of exposure during preconception. However, results from a recent simulation study have shown live birth bias may have a negative bias which would mean results would likely be stronger than was what observed within this study if present [415]. Conclusions This study provides compelling evidence that early to mid-pregnancy exposure to ambient air pollution, particularly PM2.5, influences fetal growth in utero. Our results add to the growing literature that exposure to ambient air pollution during the susceptible prenatal period influences fetal health and development, which ultimately may affect health later in life. Air pollution is a modifiable environmental exposure and continued efforts to reduce air pollution are needed to protect vulnerable populations. 135 Supplemental Material Supplemental Table 1. Demographics of 863 participants within the MADRES Study Characteristic Mean (SD) / N (%) Race/Ethnicity (N=811) Hispanic Non-Hispanic Black Non-Hispanic White Non-Hispanic Other 637 (78.5%) 102 (12.6%) 44 (5.4%) 28 (3.5%) Household Income (N=804) <$50,000 >$50,000 Reported “Don’t Know” 454 (56.5%) 78 (9.7%) 272 (33.8%) Education (N=787) High School Diploma or Less Some College/Technical School 459 (58.3%) 328 (41.7%) Age at recruitment (years) 28.3 (6.1) Pre-pregnancy BMI (N=768) 28.8 (6.8) Any Prenatal Smoking 17 (2.0%) 136 Supplemental Figure 1. Spearman Correlations of Pollutants Weeks 1-32 of Gestation 137 Supplemental Figure 2. Pearson Correlations of Fetal Growth Outcomes with Infant Birthweight Note: HC= Head Circumference, AC=abdominal circumference, BPD=Biparietal Diameter, EFW=Estimated Fetal Weight 138 Supplemental Figure 3. DLM Model Results for PM 10 and Fetal Abdominal Circumference Stratified by Fetal Sex Adjusted for maternal age at time of study recruitment, maternal race and ethnicity, maternal education level, household income, parity, pre-pregnancy body mass index (BMI), gestational age at time of ultrasound scan, lag specific average temperature, season of ultrasound, ultrasound technician, and recruitment site. 139 PM 10 PM 2.5 PM 2.5 Supplemental Figure 4. Results of DLM Models After Additionally Adjusting for Chronic/Gestational Diabetes and Hypertension Adjusted for maternal age at time of study recruitment, maternal race and ethnicity, maternal education level, household income, parity, pre-pregnancy body mass index (BMI), sex of the fetus, gestational age at time of ultrasound scan, lag specific average temperature, season of ultrasound, ultrasound technician, recruitment site, and maternal health complications (chronic/gestational diabetes and hypertension). 140 PM 10 PM 2.5 PM 2.5 Supplemental Figure 5. Results of DLM Models After Removing Participants Who Reported Any Personal Smoking (N=7) Adjusted for maternal age at time of study recruitment, maternal race and ethnicity, maternal education level, household income, parity, pre-pregnancy body mass index (BMI), sex of the fetus, gestational age at time of ultrasound scan, lag specific average temperature, season of ultrasound, ultrasound technician, and recruitment site. 141 CHAPTER SEVEN Summary Gestation represents a potentially susceptible period to environmental exposures that may have lasting impacts on fetal development and later childhood and adulthood diseases [416, 417]. Strong evidence exists indicating that environmental exposures to chemical contaminants differ by race/ethnicity, as does the likelihood of adverse pregnancy outcomes such as having an infant born with low birthweight (<2,500 grams) [418, 419]. Populations facing health disparities tend to live in communities that house multiple sources of pollution due to the proximity to major traffic sources and industrial facilities. However, beyond higher exposure, these communities may be more vulnerable to health effects from these pollutants due to a compounded challenge of also facing inequities related to shortages in health care access, economic disadvantages, social stressors, and potential language barriers [420, 421]. Previous research assessing the influence of both per-and polyfluoroalkyl substances (PFAS) and ambient air pollution exposure during pregnancy on fetal growth has predominately been evaluated using infant birthweight. Additionally, most of the literature has been conducted among homogenous and more affluent populations. Therefore, my dissertation aimed to examine the association of two prenatal environmental exposures independently on fetal growth, with a focus on growth measured in utero, within the Maternal And Developmental Risks from Environmental and Social stressors (MADRES) pregnancy cohort. The MADRES cohort is comprised of a structurally marginalized population facing health disparities in urban Los Angeles, California. There were several intended contributions to the literature this work hoped to accomplish. First, PFAS levels were measured, and demographic differences were characterized, within a 142 predominately low-income Hispanic population that has not been well studied (Study 1). Additionally, the influence of prenatal PFAS exposure and infant birthweight was explored while ensuring gestational age did not mediate this association. Notably, this study addressed limitations within previous literature of PFAS and infant birthweight by including pivotal covariates, using multipollutant models, assessing replacement and legacy compounds (i.e. PFOA, PFOS), and conducting sensitivity analyses that have been suggested as possible biases that have previously been unaddressed (Study 1). Beyond birth outcomes, this dissertation also sought to assess how prenatal PFAS exposure influences longitudinal fetal growth trajectories in utero, as current literature on this association is limited (Study 2). As previously discussed, health disparities populations may be more vulnerable to health effects from contaminants due to structural inequities and stressors [5, 6]. For these reasons, this study also explored how maternal perceived stress experienced during pregnancy modified exposure-fetal growth associations, which has not been well studied (Study 2). Lastly, this work took an innovative approach to assessing how prenatal ambient air pollution influences growth in utero by using distributed lag modeling (DLM) techniques that allowed susceptible windows of exposure to be identified with excellent temporal accuracy (Study 3). This dissertation’s main findings were the following: 1. Prenatal PFAS concentration levels of 14 analytes measured in maternal blood serum in a subset of MADRES participants had low detection rates overall and generally lower median concentrations compared to other pregnancy cohorts. Within the five analytes that had at least 50% of samples detected, we found significant differences by maternal race, household income, education level, nativity, parity, and dietary fish intake during pregnancy. 143 2. Prenatal perfluorooctanoic acid (PFOA) exposure was significantly associated with decreased infant birthweight and results were consistent with birthweight for gestational age z-scores that were sex or parity specific. Similar conclusions remained when using a multipollutant framework. We found no association with gestational age at birth, which suggests that PFOA exposure does not influence infant birthweight through a mediating effect of earlier gestational age at birth. We found no evidence for effect modification by infant sex. 3. Prenatal PFOA exposure was significantly associated with decreased fetal head biometry growth trajectories in utero reflected by head circumference and biparietal diameter. We found evidence for this association to be modified by perceived stress experienced during pregnancy. Effect estimates of PFOA on fetal growth parameters were stronger and only significant in participants with higher perceived stress. 4. Several critical windows of exposure to prenatal air pollution exposure on in utero fetal growth were identified through DLMs. Particulate matter with an aerodynamic diameter of 2.5 μm or less (PM2.5) during weeks four to sixteen of gestation was associated with lower estimated fetal weight measured in the third trimester of pregnancy. PM2.5 exposure during weeks one to twenty-three weeks of pregnancy was significantly associated with decreased fetal abdominal circumference. Particulate matter with an aerodynamic diameter of 10 μm or less (PM10) exposure during weeks six to fifteen of pregnancy was associated with reduced third trimester fetal abdominal circumference. We found evidence that associations between prenatal PM10 exposure and abdominal circumference growth may be modified by fetal sex, with female fetuses being more affected. The impact of preconception air pollution exposure (3 months 144 prior to conception) on fetal growth was assessed, but no significant windows were identified. Conclusions and Implications This dissertation found evidence that prenatal exposure to both PFAS and ambient air pollution contributes to declines in fetal growth across gestation within a health disparities population. Specifically, PFOA was associated with fetal growth in two ways: 1) by decreasing infant birthweight and 2) by impacting in utero fetal growth trajectories with declines in head circumference and biparietal diameter. Previous reports have estimated that low birthweight attributed to PFOA exposure from 2003-2014 cost the US $13.7 billion dollars from hospitalizations and lost economic productivity [422]. PFOA began to be phased out of US production beginning in the early 2000s, subsequently leading to Teflon (non-stick) pans to not include this particular compound. A study in 2014, found that since the phase-out, low birthweight exclusively attributed to PFOA exposure through Teflon declined by at least 10,000 cases a year [422]. Despite being phased out of production, adverse influences on fetal growth are still being observed among MADRES participants with prenatal PFOA levels measured in blood serum from late 2015 to early 2019. The recent measurement paints a more accurate picture of current exposure levels within the population and also sheds light on how immigration to the US may change exposure patterns since significant differences were observed based on nativity. Our findings have important implications on prenatal public health. First, there are lingering effects of previous PFOA exposure or recent PFOA exposure that remains relevant in recent years and influences fetal growth. New exposure likely occurs through dietary ingestion due to environmental persistence or the use of imported products. A large strength of both Study 145 1 and Study 2 is that prenatal PFAS were directly measured from biospecimens, which allowed an accurate exposure profile to be painted by capturing the internal dose within MADRES participants. However, we were unable to determine the exact exposure pathways that led to higher levels in some participants compared to others. With PFAS having multiple exposure pathways, interventions should be centered on governmental regulations as these results add to literature supporting interventions, education, and data needed for policy makers. The expected impact of this study is to provide further evidence to promote reductions in PFAS exposure, including by remediation efforts and addressing groundwater that remains contaminated with PFOA. Currently, beyond PFOA and PFOS no longer being able to be manufactured within the US, additional regulations are only at the state level. Regulations are predominately centered on levels found in drinking water, with only a handful of states participating [423]. Specifically, California has one of the smallest allowable concentrations of 5.1 parts per trillion (ppt) in drinking water for PFOA [8], PFNA, PFOS, and PFOA are listed as reproductive/developmental toxicants on the Proposition 65 chemical list [424] and recent legislation was passed through Assembly Bill 1200 (AB-1200) for PFAS to not be present in food packaging as well as for cookware to properly disclose which chemicals are used [425]. Additionally, this research potentially elucidates the importance of education and disclosure to pregnant persons, or those trying to get pregnant soon, on avoiding future potentially modifiable PFAS exposure. Possible examples include reducing use of products with high levels of PFAS (such as in certain shampoos, fast food packaging, or cookware), avoiding high quantities of seafood, or drinking filtered water over tap water with possible governmental aid to make this financially feasible in lower-income groups. 146 This current work adds to the literature that prenatal PFOA negatively impacts fetal growth and included several under-researched analytes; however, there are nearly 4,700 PFAS compounds. Our results warrant future work on these compounds that share similar chemical structures as many of these analytes have no restrictions, human exposure levels have not been quantified, and they have not been examined for their downstream effects on human health. Lastly, this dissertation provided novel evidence that PFAS exposure and perceived stress experienced in pregnancy may interact to adversely impact fetal growth. This should be considered clinically for the importance of reducing stress throughout gestation, especially since an individual’s previous PFAS exposure is not modifiable. Our innovative results for prenatal air pollution exposure on in utero fetal growth indicated that pregnant persons are most susceptible to particulate matter exposure during early to mid- pregnancy. These results are considered to have excellent temporal resolution for outdoor air pollution exposure based on residential locations. This resolution comes from pollutant concentrations being estimated at an hourly level to create daily and weekly averages and participant residential locations being ascertained with higher confidence at the daily level, but additionally by using Distributed Lag Models (DLM), which properly captures the variability of the exposure across gestation to identify finer windows of critical exposure (weekly). Ambient air pollution exposure experienced prior to conception and across gestation was quantified in this study using inverse-distance-squared weighted interpolation of ambient pollution data from EPA air quality monitors. The National Ambient Air Quality Standards, or NAAQS, have been set in large part based on a critical review of previous literature documenting the adverse health risks of exposure to outdoor air pollution [426]. The median PM2.5 and PM10 exposure levels across gestation within this study were below current regulatory standards set by 147 NAAQS [427], which indicates that air pollution meeting the current annual standards may still be negatively impacting fetal growth. Using data from 2010, 3.3% of preterm births were attributed to prenatal PM2.5 exposure, leading to an economic burden of over five billion dollars [428]. Even among infants born full term, the economic burden of decreased growth due to air pollution is likely much higher since fetal head growth is a strong predictor of childhood and adulthood cognitive outcomes, including intelligence quotient [429]. Although over recent decades ambient air pollution levels have decreased in many urban areas, including in Los Angeles, air pollution is a modifiable environmental exposure and continued efforts to reduce air pollution especially in marginalized communities are needed to protect vulnerable populations. When NAAQS are reassessed to determine new health protective standards based on literature documenting health risks, we hope that this work adds to the body of evidence showing that current levels of air pollution are still contributing to adverse health effects in children. Future Studies I have several suggested future directions for subsequent studies based on the analyses and results from this dissertation. Originally, this dissertation had intended to use mixture methods approaches to assess the association of prenatal PFAS on infant birthweight and in utero fetal growth due to humans being exposed to a myriad of pollutants concurrently. While we were pleased to learn that the subset of MADRES participants with PFAS concentrations were lower than expected, with many analytes having low or no detection, we were limited in our analytic approaches as the majority of measured analytes were excluded. With those remaining, more than half needed to be modeled categorically (detected/non-detected) given a high frequency of non- 148 detects. In turn, the number of variables or exposures and thus assumptions required to properly conduct principal component analysis (PCA) or Bayesian kernel machine regression (BKMR) were not met. Although we did incorporate a multipollutant framework in Study 1, future studies with higher detection frequencies should use mixture methods as an important next step. This is especially imperative to further confirm the potential synergistic role of maternal perceived stress with PFAS exposure mixtures on in utero fetal growth. Given evidence that prenatal PFAS exposure, particularly PFOA, influences in utero fetal growth and birthweight, a natural next step would be to identify what the major sources and pathways of PFOA exposure are in this population or similar health disparities groups and to further explore how early life PFAS exposure influences infant and early childhood growth trajectories, and whether breastfeeding modifies this relationship. Research has shown PFAS levels to accumulate in infants by 20-30% per month they are breastfed [430]. I would like to further this work by assessing prenatal PFAS levels, as well as PFAS levels found in breast milk, with infant growth trajectories and investigate whether breastfeeding status modifies these associations. I would also like to broaden infant growth trajectories beyond weight gain and assess how prenatal PFAS exposure influences adiposity by specifically looking at body composition fat levels. Previous literature has suggested that PFAS acts as an obesogen [431, 432]. The MADRES study has access to novel technology using quantitative nuclear magnetic resonance (QMR) known as EchoMRI TM Adolescent Humans Body Composition Analyzer (EchoMRI-AH, Echo Medical System, Houston, TX). The EchoMRI-AH assesses whole body composition in less than 3.2 minutes by counting hydrogen atoms using a low strength magnetic field to provide measurements of fat mass, lean mass, free water and total body water in infants and toddlers up to 80 kg. A key 149 strength of this approach is that it emits no radiation and does not require the child to be immobilized, so it is safe and feasible to use in young children. The technology provides automated values of each measurement, which aggregates an average from three sequential scans. MADRES assesses body composition using the EchoMRI-AH at several times across infancy and childhood. I propose several next steps for my work focusing on prenatal air pollution exposure and in utero fetal growth. First, there are statistical methods that incorporate mixtures methods with lagged exposures that vary over time [433-437]. I would like to further explore the sensitive windows identified for particulate matter exposure on fetal growth while properly accounting for the ambient air pollutant mixture concurrently exposed to mothers and their developing fetus. I would also be interested in incorporating PFAS, particularly PFOA, into this mixtures analysis. Since both environmental pollutants independently show adverse influences on in utero fetal development, I believe it is important to look at their combined effect in this framework. Recent literature has shown inhalation of indoor air as an important exposure route for PFAS to humans [438]. Beyond mixtures of outdoor air pollutants, I would also be interested in specifically looking at the complex mixture that makes up particulate matter on fetal growth by looking at personal monitoring data. Particulate matter does not have a uniform chemical composition and is made up of organic compounds, metals, dust, acids and other components. The outdoor mixture found within Los Angeles might be specific to the region, and future work could assess which components of PM2.5 and PM10 have the strongest influences on growth in utero. Lastly, I believe it is also important to investigate whether maternal exposure to psychosocial stress experienced during pregnancy serves as a modifier between prenatal air pollution and in utero fetal growth, as we observed within the PFAS analyses. 150 Conclusions This dissertation found evidence that exposure to environmental contaminants during pregnancy is associated with reduced fetal growth across gestation and higher likelihood of adverse birth outcomes. Additionally, results also suggested that vulnerability to exposures could be compounded by stress. We also identified critical windows of ambient air pollution exposure across gestation that were associated with altered fetal growth trajectories. Further research is needed to assess the influence of multiple environmental contaminants concurrently on in utero fetal growth. 151 REFERENCES 1. Delpierre, C., et al., [DOHaD: epidemiological researches]. Med Sci (Paris), 2016. 32(1): p. 21-6. 2. Thomas, D.C., Statistical methods in environmental epidemiology. 2009, Oxford: Oxford University Press. 3. Kramer, M.S., The epidemiology of adverse pregnancy outcomes: an overview. The Journal of nutrition, 2003. 133(5): p. 1592S-1596S. 4. Stillerman, K.P., et al., Environmental exposures and adverse pregnancy outcomes: a review of the science. Reprod Sci, 2008. 15(7): p. 631-50. 5. Mattison, D.R., Environmental exposures and development. Curr Opin Pediatr, 2010. 22(2): p. 208-18. 6. EPA, U., TSCA Chemical Substance Inventory. In: US EPA [Internet]. 2015. 7. Varshavsky, J., et al., Heightened susceptibility: A review of how pregnancy and chemical exposures influence maternal health. Reprod Toxicol, 2020. 92: p. 14-56. 8. Heindel, J.J. and L.N. Vandenberg, Developmental origins of health and disease: a paradigm for understanding disease cause and prevention. Curr Opin Pediatr, 2015. 27(2): p. 248-53. 9. Autrup, H., Transplacental transfer of genotoxins and transplacental carcinogenesis. Environ Health Perspect, 1993. 101 Suppl 2(Suppl 2): p. 33-8. 10. Grippo, A., et al., Air pollution exposure during pregnancy and spontaneous abortion and stillbirth. Rev Environ Health, 2018. 33(3): p. 247-264. 11. Mamsen, L.S., et al., Concentrations of perfluoroalkyl substances (PFASs) in human embryonic and fetal organs from first, second, and third trimester pregnancies. Environ Int, 2019. 124: p. 482-492. 12. Cutland, C.L., et al., Low birth weight: Case definition & guidelines for data collection, analysis, and presentation of maternal immunization safety data. Vaccine, 2017. 35(48 Pt A): p. 6492-6500. 13. Martin, J.A., et al., Births: Final Data for 2019. Natl Vital Stat Rep, 2021. 70(2): p. 1-51. 14. Barker, D.J., et al., Type 2 (non-insulin-dependent) diabetes mellitus, hypertension and hyperlipidaemia (syndrome X): relation to reduced fetal growth. Diabetologia, 1993. 36(1): p. 62-7. 152 15. Evensen, E., et al., The relation between birthweight, childhood body mass index, and overweight and obesity in late adolescence: a longitudinal cohort study from Norway, The Tromsø Study, Fit Futures. BMJ open, 2017. 7(6). 16. Jornayvaz, F.R., et al., Low birth weight leads to obesity, diabetes and increased leptin levels in adults: the CoLaus study. Cardiovascular diabetology, 2016. 15(1): p. 1-10. 17. Wang, S.-F., et al., Birth weight and risk of coronary heart disease in adults: a meta- analysis of prospective cohort studies. Journal of developmental origins of health and disease, 2014. 5(6): p. 408-419. 18. Luu, T.M., M.O.R. Mian, and A.M. Nuyt, Long-term impact of preterm birth: neurodevelopmental and physical health outcomes. Clinics in perinatology, 2017. 44(2): p. 305-314. 19. Riley, W.J., Health disparities: gaps in access, quality and affordability of medical care. Transactions of the American Clinical and Climatological Association, 2012. 123: p. 167. 20. Tanya Nagahawatte, N. and R.L. Goldenberg, Poverty, maternal health, and adverse pregnancy outcomes. Annals of the New York Academy of Sciences, 2008. 1136(1): p. 80-85. 21. Mohai, P., D. Pellow, and J.T. Roberts, Environmental Justice. Annual Review of Environment and Resources, 2009. 34(1): p. 405-430. 22. Zwickl, K., M. Ash, and J.K. Boyce, Regional variation in environmental inequality: Industrial air toxics exposure in US cities. Ecological Economics, 2014. 107: p. 494-509. 23. Mohai, P. and R. Saha, Racial inequality in the distribution of hazardous waste: A national-level reassessment. Social problems, 2007. 54(3): p. 343-370. 24. Cushing, L., et al., Racial/Ethnic Disparities in Cumulative Environmental Health Impacts in California: Evidence From a Statewide Environmental Justice Screening Tool (CalEnviroScreen 1.1). Am J Public Health, 2015. 105(11): p. 2341-8. 25. Alkandari, F., et al., Fetal ultrasound measurements and associations with postnatal outcomes in infancy and childhood: a systematic review of an emerging literature. J Epidemiol Community Health, 2015. 69(1): p. 41-48. 26. Gale, C.R., et al., The influence of head growth in fetal life, infancy, and childhood on intelligence at the ages of 4 and 8 years. Pediatrics, 2006. 118(4): p. 1486-92. 27. Vesterinen, H.M., et al., Cumulative effects of prenatal-exposure to exogenous chemicals and psychosocial stress on fetal growth: Systematic-review of the human and animal evidence. PLoS One, 2017. 12(7): p. e0176331. 153 28.. Sinclair, G.M., S.M. Long, and O.A. Jones, What are the effects of PFAS exposure at environmentally relevant concentrations? Chemosphere, 2020: p. 127340. 29. Ding, G. and W.J. Peijnenburg, Physicochemical properties and aquatic toxicity of poly- and perfluorinated compounds. Critical reviews in environmental science and technology, 2013. 43(6): p. 598-678. 30. Smart, B., et al., Organofluorine Chemistry. Principles and Commercial Applications, 1994: p. 57. 31. Ateia, M., et al., The overlooked short-and ultrashort-chain poly-and perfluorinated substances: A review. Chemosphere, 2019. 220: p. 866-882. 32. Coggan, T.L., et al., A single analytical method for the determination of 53 legacy and emerging per-and polyfluoroalkyl substances (PFAS) in aqueous matrices. Analytical and bioanalytical chemistry, 2019. 411(16): p. 3507-3520. 33. Coggan, T.L., et al., An investigation into per-and polyfluoroalkyl substances (PFAS) in nineteen Australian wastewater treatment plants (WWTPs). Heliyon, 2019. 5(8): p. e02316. 34. Li, Y., et al., Half-lives of PFOS, PFHxS and PFOA after end of exposure to contaminated drinking water. Occup Environ Med, 2018. 75(1): p. 46-51. 35. Olsen, G.W., et al., Half-life of serum elimination of perfluorooctanesulfonate,perfluorohexanesulfonate, and perfluorooctanoate in retired fluorochemical production workers. Environ Health Perspect, 2007. 115(9): p. 1298-305. 36. Pelch, K.E., et al., PFAS health effects database: Protocol for a systematic evidence map. Environment international, 2019. 130: p. 104851. 37. Xu, Y., et al., Serum Half-Lives for Short- and Long-Chain Perfluoroalkyl Acids after Ceasing Exposure from Drinking Water Contaminated by Firefighting Foam. Environ Health Perspect, 2020. 128(7): p. 77004. 38. Trojanowicz, M. and M. Koc, Recent developments in methods for analysis of perfluorinated persistent pollutants. Mikrochim Acta, 2013. 180(11): p. 957-971. 39. Fair, P.A., et al., Perfluoroalkyl substances (PFASs) in edible fish species from Charleston Harbor and tributaries, South Carolina, United States: Exposure and risk assessment. Environ Res, 2019. 171: p. 266-277. 40. Hoffman, K., et al., Private drinking water wells as a source of exposure to perfluorooctanoic acid (PFOA) in communities surrounding a fluoropolymer production facility. Environ Health Perspect, 2011. 119(1): p. 92-7. 154 41. Schaider, L.A., et al., Fluorinated Compounds in U.S. Fast Food Packaging. Environ Sci Technol Lett, 2017. 4(3): p. 105-111. 42. Tanner, E.M., et al., Occupational exposure to perfluoroalkyl substances and serum levels of perfluorooctanesulfonic acid (PFOS) and perfluorooctanoic acid (PFOA) in an aging population from upstate New York: a retrospective cohort study. Int Arch Occup Environ Health, 2018. 91(2): p. 145-154. 43. Tittlemier, S.A., et al., Dietary exposure of Canadians to perfluorinated carboxylates and perfluorooctane sulfonate via consumption of meat, fish, fast foods, and food items prepared in their packaging. J Agric Food Chem, 2007. 55(8): p. 3203-10. 44. Calafat, A.M., et al., Polyfluoroalkyl chemicals in the U.S. population: data from the National Health and Nutrition Examination Survey (NHANES) 2003-2004 and comparisons with NHANES 1999-2000. Environ Health Perspect, 2007. 115(11): p. 1596- 602. 45. Bassler, J., et al., Environmental perfluoroalkyl acid exposures are associated with liver disease characterized by apoptosis and altered serum adipocytokines. Environ Pollut, 2019. 247: p. 1055-1063. 46. Chen, J., et al., Chronic PFOS Exposure Disrupts Thyroid Structure and Function in Zebrafish. Bull Environ Contam Toxicol, 2018. 101(1): p. 75-79. 47. Chen, X., et al., Perfluorooctanesulfonate induces neuroinflammation through the secretion of TNF-α mediated by the JAK2/STAT3 pathway. Neurotoxicology, 2018. 66: p. 32-42. 48. Chou, H.C., et al., From the Cover: l-Carnitine via PPARγ- and Sirt1-Dependent Mechanisms Attenuates Epithelial-Mesenchymal Transition and Renal Fibrosis Caused by Perfluorooctanesulfonate. Toxicol Sci, 2017. 160(2): p. 217-229. 49. Han, R., et al., Effect of perfluorooctane sulphonate-induced Kupffer cell activation on hepatocyte proliferation through the NF-κB/TNF-α/IL-6-dependent pathway. Chemosphere, 2018. 200: p. 283-294. 50. Mao, Z., et al., Perfluorooctane sulfonate induces apoptosis in lung cancer A549 cells through reactive oxygen species-mediated mitochondrion-dependent pathway. J Appl Toxicol, 2013. 33(11): p. 1268-76. 51. Soloff, A.C., et al., Environmental perfluorooctane sulfonate exposure drives T cell activation in bottlenose dolphins. J Appl Toxicol, 2017. 37(9): p. 1108-1116. 52. Steenland, K., T. Fletcher, and D.A. Savitz, Epidemiologic evidence on the health effects of perfluorooctanoic acid (PFOA). Environ Health Perspect, 2010. 118(8): p. 1100-8. 155 53. Tang, L.L., et al., Mitochondrial toxicity of perfluorooctane sulfonate in mouse embryonic stem cell-derived cardiomyocytes. Toxicology, 2017. 382: p. 108-116. 54. Zeng, Z., et al., Assessing the human health risks of perfluorooctane sulfonate by in vivo and in vitro studies. Environ Int, 2019. 126: p. 598-610. 55. EPA, U., 2010/2015 PFOA Stewardship Program. 2006, United States Environmental Protection Agency Washington, DC. 56. OEHHA, C., The Proposition 65 List. 2021, OEHHA. https://oehha. ca. gov/proposition- 65/proposition-65-list. 57. Canada, G.o., Canada gazette, part I, volume 153, number 3: Government notices. 2019. 58. ECHA. Perfluroalkyl chemicals (PFAS). [cited 2022; Available from: https://echa.europa.eu/hot-topics/perfluoroalkyl-chemicals-pfas. 59. EPA, U., Basic Information on PFAS. 2016. 60. Information, C.L. Assembly Bill No. 1200. 2021 [cited 2022; Available from: https://leginfo.legislature.ca.gov/faces/billNavClient.xhtml?bill_id=202120220AB1200. 61. Barbarossa, A., et al., Perfluoroalkyl substances in human milk: a first survey in Italy. Environment international, 2013. 51: p. 27-30. 62. Kato, K., et al., Changes in serum concentrations of maternal poly- and perfluoroalkyl substances over the course of pregnancy and predictors of exposure in a multiethnic cohort of Cincinnati, Ohio pregnant women during 2003-2006. Environ Sci Technol, 2014. 48(16): p. 9600-8. 63. Mondal, D., et al., Breastfeeding: a potential excretion route for mothers and implications for infant exposure to perfluoroalkyl acids. Environ Health Perspect, 2014. 122(2): p. 187-92. 64. Tao, L., et al., Perfluorinated compounds in human milk from Massachusetts, U.S.A. Environ Sci Technol, 2008. 42(8): p. 3096-101. 65. Thomsen, C., et al., Changes in concentrations of perfluorinated compounds, polybrominated diphenyl ethers, and polychlorinated biphenyls in Norwegian breast-milk during twelve months of lactation. Environ Sci Technol, 2010. 44(24): p. 9550-6. 66. Winkens, K., et al., Early life exposure to per-and polyfluoroalkyl substances (PFASs): A critical review. Emerging Contaminants, 2017. 3(2): p. 55-68. 156 67. Boronow, K.E., et al., Serum concentrations of PFASs and exposure-related behaviors in African American and non-Hispanic white women. Journal of exposure science & environmental epidemiology, 2019. 29(2): p. 206-217. 68. Brantsæter, A.L., et al., Determinants of plasma concentrations of perfluoroalkyl substances in pregnant Norwegian women. Environ Int, 2013. 54: p. 74-84. 69. Tsai, M.S., et al., Determinants and Temporal Trends of Perfluoroalkyl Substances in Pregnant Women: The Hokkaido Study on Environment and Children's Health. Int J Environ Res Public Health, 2018. 15(5). 70. Qi, W., et al., Per- and Polyfluoroalkyl Substances and Obesity, Type 2 Diabetes and Non-alcoholic Fatty Liver Disease: A Review of Epidemiologic Findings. Toxicol Environ Chem, 2020. 102(1-4): p. 1-36. 71. Blazer, S., et al., Maternal hypothyroidism may affect fetal growth and neonatal thyroid function. Obstet Gynecol, 2003. 102(2): p. 232-41. 72. Bach, C.C., et al., Perfluoroalkyl and polyfluoroalkyl substances and human fetal growth: a systematic review. Crit Rev Toxicol, 2015. 45(1): p. 53-67. 73. Du, G., et al., Perfluorooctane sulfonate (PFOS) affects hormone receptor activity, steroidogenesis, and expression of endocrine-related genes in vitro and in vivo. Environ Toxicol Chem, 2013. 32(2): p. 353-60. 74. Lau, C., et al., Exposure to perfluorooctane sulfonate during pregnancy in rat and mouse. II: postnatal evaluation. Toxicol Sci, 2003. 74(2): p. 382-92. 75. Thibodeaux, J.R., et al., Exposure to perfluorooctane sulfonate during pregnancy in rat and mouse. I: maternal and prenatal evaluations. Toxicol Sci, 2003. 74(2): p. 369-81. 76. Kaijser, M., et al., Maternal pregnancy estriol levels in relation to anamnestic and fetal anthropometric data. Epidemiology, 2000. 11(3): p. 315-9. 77. Bach, C.C., et al., Perfluoroalkyl and polyfluoroalkyl substances and measures of human fertility: a systematic review. Crit Rev Toxicol, 2016. 46(9): p. 735-55. 78. Benninghoff, A.D., et al., Estrogen-like activity of perfluoroalkyl acids in vivo and interaction with human and rainbow trout estrogen receptors in vitro. Toxicol Sci, 2011. 120(1): p. 42-58. 79. Henry, N.D. and P.A. Fair, Comparison of in vitro cytotoxicity, estrogenicity and anti- estrogenicity of triclosan, perfluorooctane sulfonate and perfluorooctanoic acid. J Appl Toxicol, 2013. 33(4): p. 265-72. 157 80. Kjeldsen, L.S. and E.C. Bonefeld-Jørgensen, Perfluorinated compounds affect the function of sex hormone receptors. Environ Sci Pollut Res Int, 2013. 20(11): p. 8031-44. 81. Szilagyi, J.T., V. Avula, and R.C. Fry, Perfluoroalkyl Substances (PFAS) and Their Effects on the Placenta, Pregnancy, and Child Development: a Potential Mechanistic Role for Placental Peroxisome Proliferator-Activated Receptors (PPARs). Curr Environ Health Rep, 2020. 7(3): p. 222-230. 82. Blake, B.E. and S.E. Fenton, Early life exposure to per- and polyfluoroalkyl substances (PFAS) and latent health outcomes: A review including the placenta as a target tissue and possible driver of peri- and postnatal effects. Toxicology, 2020. 443: p. 152565. 83. Suh, C.H., et al., Perfluorooctanoic acid-induced inhibition of placental prolactin-family hormone and fetal growth retardation in mice. Mol Cell Endocrinol, 2011. 337(1-2): p. 7-15. 84. Glazier, J.D., et al., Association between the activity of the system A amino acid transporter in the microvillous plasma membrane of the human placenta and severity of fetal compromise in intrauterine growth restriction. Pediatr Res, 1997. 42(4): p. 514-9. 85. Mahendran, D., et al., Amino acid (system A) transporter activity in microvillous membrane vesicles from the placentas of appropriate and small for gestational age babies. Pediatr Res, 1993. 34(5): p. 661-5. 86. Mamsen, L.S., et al., Concentrations of perfluoroalkyl substances (PFASs) in human embryonic and fetal organs from first, second, and third trimester pregnancies. Environ Int, 2019. 124: p. 482-492. 87. Chen, L., et al., Prenatal exposure to perfluoroalkyl and polyfluoroalkyl substances and birth outcomes: A longitudinal cohort with repeated measurements. Chemosphere, 2021. 267: p. 128899. 88. Koskela, A., et al., Perfluoroalkyl substances in human bone: concentrations in bones and effects on bone cell differentiation. Sci Rep, 2017. 7(1): p. 6841. 89. DeWitt, J.C., et al., Immunotoxicity of perfluorinated compounds: recent developments. Toxicol Pathol, 2012. 40(2): p. 300-11. 90. Grandjean, P., et al., Serum vaccine antibody concentrations in children exposed to perfluorinated compounds. Jama, 2012. 307(4): p. 391-7. 91. Zheng, T., et al., Effects of Environmental Exposures on Fetal and Childhood Growth Trajectories. Ann Glob Health, 2016. 82(1): p. 41-99. 158 92. Abbott, B.D., et al., Perfluorooctanoic acid induced developmental toxicity in the mouse is dependent on expression of peroxisome proliferator activated receptor-alpha. Toxicol Sci, 2007. 98(2): p. 571-81. 93. Lau, C., et al., Effects of perfluorooctanoic acid exposure during pregnancy in the mouse. Toxicol Sci, 2006. 90(2): p. 510-8. 94. Wolf, C.J., et al., Developmental toxicity of perfluorooctanoic acid in the CD-1 mouse after cross-foster and restricted gestational exposures. Toxicol Sci, 2007. 95(2): p. 462- 73. 95. Koustas, E., et al., The Navigation Guide—evidence-based medicine meets environmental health: systematic review of nonhuman evidence for PFOA effects on fetal growth. Environmental health perspectives, 2014. 122(10): p. 1015-1027. 96. Negri, E., et al., Exposure to PFOA and PFOS and fetal growth: a critical merging of toxicological and epidemiological data. Crit Rev Toxicol, 2017. 47(6): p. 482-508. 97. Olsen, G.W., J.L. Butenhoff, and L.R. Zobel, Perfluoroalkyl chemicals and human fetal development: an epidemiologic review with clinical and toxicological perspectives. Reproductive Toxicology, 2009. 27(3-4): p. 212-230. 98. Apelberg, B.J., et al., Cord serum concentrations of perfluorooctane sulfonate (PFOS) and perfluorooctanoate (PFOA) in relation to weight and size at birth. Environmental health perspectives, 2007. 115(11): p. 1670-1676. 99. Fei, C., et al., Perfluorinated chemicals and fetal growth: a study within the Danish National Birth Cohort. Environ Health Perspect, 2007. 115(11): p. 1677-82. 100. Maisonet, M., et al., Maternal concentrations of polyfluoroalkyl compounds during pregnancy and fetal and postnatal growth in British girls. Environ Health Perspect, 2012. 120(10): p. 1432-7. 101. Starling, A.P., et al., Perfluoroalkyl substances during pregnancy and offspring weight and adiposity at birth: examining mediation by maternal fasting glucose in the healthy start study. Environmental health perspectives, 2017. 125(6): p. 067016. 102. Wikström, S., et al., Maternal serum levels of perfluoroalkyl substances in early pregnancy and offspring birth weight. Pediatr Res, 2020. 87(6): p. 1093-1099. 103. Ehresman, D.J., et al., Comparison of human whole blood, plasma, and serum matrices for the determination of perfluorooctanesulfonate (PFOS), perfluorooctanoate (PFOA), and other fluorochemicals. Environ Res, 2007. 103(2): p. 176-84. 104. Apelberg, B.J., et al., Determinants of fetal exposure to polyfluoroalkyl compounds in Baltimore, Maryland. Environ Sci Technol, 2007. 41(11): p. 3891-7. 159 105. Fromme, H., et al., Pre- and postnatal exposure to perfluorinated compounds (PFCs). Environ Sci Technol, 2010. 44(18): p. 7123-9. 106. Lee, Y.J., et al., Concentrations of perfluoroalkyl compounds in maternal and umbilical cord sera and birth outcomes in Korea. Chemosphere, 2013. 90(5): p. 1603-9. 107. Midasch, O., et al., Transplacental exposure of neonates to perfluorooctanesulfonate and perfluorooctanoate: a pilot study. Int Arch Occup Environ Health, 2007. 80(7): p. 643-8. 108. Mamsen, L.S., et al., Concentration of perfluorinated compounds and cotinine in human foetal organs, placenta, and maternal plasma. Sci Total Environ, 2017. 596-597: p. 97- 105. 109. Johnson, P.I., et al., The Navigation Guide - evidence-based medicine meets environmental health: systematic review of human evidence for PFOA effects on fetal growth. Environ Health Perspect, 2014. 122(10): p. 1028-39. 110. Steenland, K., V. Barry, and D. Savitz, Serum Perfluorooctanoic Acid and Birthweight: An Updated Meta-analysis With Bias Analysis. Epidemiology, 2018. 29(6): p. 765-776. 111. Darrow, L.A., C.R. Stein, and K. Steenland, Serum perfluorooctanoic acid and perfluorooctane sulfonate concentrations in relation to birth outcomes in the Mid-Ohio Valley, 2005-2010. Environ Health Perspect, 2013. 121(10): p. 1207-13. 112. Wu, K., et al., Association between maternal exposure to perfluorooctanoic acid (PFOA) from electronic waste recycling and neonatal health outcomes. Environ Int, 2012. 48: p. 1-8. 113. Verner, M.A., et al., Associations of Perfluoroalkyl Substances (PFAS) with Lower Birth Weight: An Evaluation of Potential Confounding by Glomerular Filtration Rate Using a Physiologically Based Pharmacokinetic Model (PBPK). Environ Health Perspect, 2015. 123(12): p. 1317-24. 114. Brantsaeter, A., et al., Determinants of plasma concentrations of perfluoroalkyl substances in pregnant Norwegian women. Environment international, 2013. 54: p. 74- 84. 115. Savitz, D.A., et al., Relationship of perfluorooctanoic acid exposure to pregnancy outcome based on birth records in the mid-Ohio Valley. Environ Health Perspect, 2012. 120(8): p. 1201-7. 116. Salas, S.P., et al., Time course of maternal plasma volume and hormonal changes in women with preeclampsia or fetal growth restriction. Hypertension, 2006. 47(2): p. 203- 8. 160 117. Wilcox, A.J., C.R. Weinberg, and O. Basso, On the pitfalls of adjusting for gestational age at birth. Am J Epidemiol, 2011. 174(9): p. 1062-8. 118. Chen, M.H., et al., Perfluorinated compounds in umbilical cord blood and adverse birth outcomes. PLoS One, 2012. 7(8): p. e42474. 119. Hamm, M.P., et al., Maternal exposure to perfluorinated acids and fetal growth. J Expo Sci Environ Epidemiol, 2010. 20(7): p. 589-97. 120. Eick, S.M., et al., Associations between prenatal maternal exposure to per-and polyfluoroalkyl substances (PFAS) and polybrominated diphenyl ethers (PBDEs) and birth outcomes among pregnant women in San Francisco. Environmental Health, 2020. 19(1): p. 1-12. 121. Costa, O., et al., First-trimester maternal concentrations of polyfluoroalkyl substances and fetal growth throughout pregnancy. Environment international, 2019. 130: p. 104830. 122. Ouidir, M., et al., Association of Maternal Exposure to Persistent Organic Pollutants in Early Pregnancy With Fetal Growth. JAMA Pediatr, 2020. 174(2): p. 149-161. 123. Grasty, R., et al., Effects of prenatal perfluorooctane sulfonate (PFOS) exposure on lung maturation in the perinatal rat. Birth Defects Research Part B: Developmental and Reproductive Toxicology, 2005. 74(5): p. 405-416. 124. Grasty, R.C., et al., Prenatal window of susceptibility to perfluorooctane sulfonate- induced neonatal mortality in the Sprague-Dawley rat. Birth Defects Res B Dev Reprod Toxicol, 2003. 68(6): p. 465-71. 125. Luebker, D.J., et al., Two-generation reproduction and cross-foster studies of perfluorooctanesulfonate (PFOS) in rats. Toxicology, 2005. 215(1-2): p. 126-48. 126. Luebker, D.J., et al., Neonatal mortality from in utero exposure to perfluorooctanesulfonate (PFOS) in Sprague-Dawley rats: dose-response, and biochemical and pharamacokinetic parameters. Toxicology, 2005. 215(1-2): p. 149-69. 127. Glynn, A., et al., Perfluorinated alkyl acids in blood serum from primiparous women in Sweden: serial sampling during pregnancy and nursing, and temporal trends 1996-2010. Environ Sci Technol, 2012. 46(16): p. 9071-9. 128. Olsen, G.W., J.L. Butenhoff, and L.R. Zobel, Perfluoroalkyl chemicals and human fetal development: an epidemiologic review with clinical and toxicological perspectives. Reprod Toxicol, 2009. 27(3-4): p. 212-230. 129. Washino, N., et al., Correlations between prenatal exposure to perfluorinated chemicals and reduced fetal growth. Environmental health perspectives, 2009. 117(4): p. 660-667. 161 130. Monroy, R., et al., Serum levels of perfluoroalkyl compounds in human maternal and umbilical cord blood samples. Environ Res, 2008. 108(1): p. 56-62. 131. Dzierlenga, M.W., L. Crawford, and M.P. Longnecker, Birth weight and perfluorooctane sulfonic acid: a random-effects meta-regression analysis. 2020, LWW. 132. Huo, X., et al., Perfluoroalkyl substances exposure in early pregnancy and preterm birth in singleton pregnancies: a prospective cohort study. Environmental Health, 2020. 19: p. 1-7. 133. Manzano-Salgado, C.B., et al., Prenatal exposure to perfluoroalkyl substances and birth outcomes in a Spanish birth cohort. Environ Int, 2017. 108: p. 278-284. 134. Stein, C.R., D.A. Savitz, and M. Dougan, Serum levels of perfluorooctanoic acid and perfluorooctane sulfonate and pregnancy outcome. Am J Epidemiol, 2009. 170(7): p. 837-46. 135. Sagiv, S.K., et al., Early-pregnancy plasma concentrations of perfluoroalkyl substances and birth outcomes in Project Viva: confounded by pregnancy hemodynamics? American journal of epidemiology, 2018. 187(4): p. 793-802. 136. Ramhøj, L., et al., Perfluorohexane Sulfonate (PFHxS) and a Mixture of Endocrine Disrupters Reduce Thyroxine Levels and Cause Antiandrogenic Effects in Rats. Toxicol Sci, 2018. 163(2): p. 579-591. 137. Tetzlaff, C.N.R., et al., Adult female rats perinatally exposed to perfluorohexane sulfonate (PFHxS) and a mixture of endocrine disruptors display increased body/fat weights without a transcriptional footprint in fat cells. Toxicology Letters, 2021. 339: p. 78-87. 138. Fenton, S.E., et al., Per- and Polyfluoroalkyl Substance Toxicity and Human Health Review: Current State of Knowledge and Strategies for Informing Future Research. Environ Toxicol Chem, 2021. 40(3): p. 606-630. 139. Meng, Q., et al., Prenatal exposure to perfluoroalkyl substances and birth outcomes; an updated analysis from the Danish National Birth Cohort. International journal of environmental research and public health, 2018. 15(9): p. 1832. 140. Marks, K.J., et al., Maternal serum concentrations of perfluoroalkyl substances and birth size in British boys. International journal of hygiene and environmental health, 2019. 222(5): p. 889-895. 141. Kashino, I., et al., Prenatal exposure to 11 perfluoroalkyl substances and fetal growth: A large-scale, prospective birth cohort study. Environment international, 2020. 136: p. 105355. 162 142. Lee, E.S., S. Han, and J.E. Oh, Association between perfluorinated compound concentrations in cord serum and birth weight using multiple regression models. Reprod Toxicol, 2016. 59: p. 53-9. 143. Workman, C.E., et al., Associations between concentrations of perfluoroalkyl substances in human plasma and maternal, infant, and home characteristics in Winnipeg, Canada. Environ Pollut, 2019. 249: p. 758-766. 144. Gardener, H., Q. Sun, and P. Grandjean, PFAS concentration during pregnancy in relation to cardiometabolic health and birth outcomes. Environ Res, 2021. 192: p. 110287. 145. Wolf, C.J., et al., Developmental effects of perfluorononanoic acid in the mouse are dependent on peroxisome proliferator-activated receptor-alpha. PPAR research, 2010. 2010. 146. GREY, B.E., R. G. ELLIS-HUTCHINGS, K. DAS, J. M. ROGERS, AND C. S. LAU. PRENATAL AND POSTNATAL EFFECTS IN RATS OF PERFLUORONONANOIC ACID EXPOSURE IN UTERO. in SOT Annual Meeting. 2008. 147. Das, K.P., et al., Developmental toxicity of perfluorononanoic acid in mice. Reprod Toxicol, 2015. 51: p. 133-44. 148. Harris, M.W. and L.S. Birnbaum, Developmental toxicity of perfluorodecanoic acid in C57BL/6N mice. Fundam Appl Toxicol, 1989. 12(3): p. 442-8. 149. Hu, J.M.Y., et al., Prenatal exposure to endocrine disrupting chemical mixtures and infant birth weight: A Bayesian analysis using kernel machine regression. Environ Res, 2021. 195: p. 110749. 150. Woods, M.M., et al., Gestational exposure to endocrine disrupting chemicals in relation to infant birth weight: a Bayesian analysis of the HOME Study. Environ Health, 2017. 16(1): p. 115. 151. Bureau, U.C., 65 and older population grows rapidly as baby boomers age. 2020. 152. Vega, W.A., M.A. Rodriguez, and E. Gruskin, Health disparities in the Latino population. Epidemiologic reviews, 2009. 31(1): p. 99-112. 153. Vesterinen, H.M., et al., Cumulative effects of prenatal-exposure to exogenous chemicals and psychosocial stress on fetal growth: Systematic-review of the human and animal evidence. PLoS One, 2017. 12(7): p. e0176331. 163 154. Dreyer, A.F., et al., Perfluoroalkyl Substance Exposure Early In Pregnancy Was Negatively Associated With Late Pregnancy Cortisone Levels. J Clin Endocrinol Metab, 2020. 105(8). 155. Eick, S.M., et al., Joint effects of prenatal exposure to per- and poly-fluoroalkyl substances and psychosocial stressors on corticotropin-releasing hormone during pregnancy. Journal of Exposure Science & Environmental Epidemiology, 2022. 32(1): p. 27-36. 156. EPA. Pollutants and Sources. 1992 [cited 2021 may 5 ]; Available from: https://www3.epa.gov/airtoxics/pollsour.html. 157. NPS. Where Does Air Pollution Come From? 2018 [cited 2021 may 5 ]; Available from: https://www.nps.gov/subjects/air/sources.htm. 158. Air Pollution. [cited 2021 May 5]; Available from: https://www.who.int/health- topics/air-pollution - tab=tab_1. 159. Suh, H.H., et al., Criteria air pollutants and toxic air pollutants. Environmental Health Perspectives, 2000. 108(suppl 4): p. 625-633. 160. Lurmann, F., E. Avol, and F. Gilliland, Emissions reduction policies and recent trends in Southern California’s ambient air quality. Journal of the Air & Waste Management Association, 2015. 65(3): p. 324-335. 161. Kim, K.-H., E. Kabir, and S. Kabir, A review on the human health impact of airborne particulate matter. Environment international, 2015. 74: p. 136-143. 162. Löndahl, J., et al., Size-resolved respiratory-tract deposition of fine and ultrafine hydrophobic and hygroscopic aerosol particles during rest and exercise. Inhal Toxicol, 2007. 19(2): p. 109-16. 163. Brook, R.D., et al., Particulate matter air pollution and cardiovascular disease: an update to the scientific statement from the American Heart Association. Circulation, 2010. 121(21): p. 2331-2378. 164. Di, Q., et al., Air pollution and mortality in the Medicare population. New England Journal of Medicine, 2017. 376(26): p. 2513-2522. 165. Dockery, D.W., et al., An association between air pollution and mortality in six U.S. cities. N Engl J Med, 1993. 329(24): p. 1753-9. 166. Krewski, D., et al., Extended follow-up and spatial analysis of the American Cancer Society study linking particulate air pollution and mortality. Res Rep Health Eff Inst, 2009(140): p. 5-114; discussion 115-36. 164 167. Bekkar, B., et al., Association of air pollution and heat exposure with preterm birth, low birth weight, and stillbirth in the US: a systematic review. JAMA network open, 2020. 3(6): p. e208243-e208243. 168. Rosa, M.J., et al., Prenatal exposure to PM2. 5 and birth weight: A pooled analysis from three North American longitudinal pregnancy cohort studies. Environment international, 2017. 107: p. 173-180. 169. Stieb, D.M., et al., Ambient air pollution, birth weight and preterm birth: a systematic review and meta-analysis. Environmental research, 2012. 117: p. 100-111. 170. Agency, U.S.E.P., America’s Children and the Environment: Third Edition. 2017: CreateSpace Independent Publishing Platform. 171. Islam, T., et al., Ozone, oxidant defense genes, and risk of asthma during adolescence. Am J Respir Crit Care Med, 2008. 177(4): p. 388-95. 172. Lin, S., et al., Chronic exposure to ambient ozone and asthma hospital admissions among children. Environ Health Perspect, 2008. 116(12): p. 1725-30. 173. Mar, T.F. and J.Q. Koenig, Relationship between visits to emergency departments for asthma and ozone exposure in greater Seattle, Washington. Ann Allergy Asthma Immunol, 2009. 103(6): p. 474-9. 174. Ha, S., et al., The effects of air pollution on adverse birth outcomes. Environ Res, 2014. 134: p. 198-204. 175. Lee, P.C., et al., First trimester exposure to ambient air pollution, pregnancy complications and adverse birth outcomes in Allegheny County, PA. Matern Child Health J, 2013. 17(3): p. 545-55. 176. Salam, M.T., et al., Birth outcomes and prenatal exposure to ozone, carbon monoxide, and particulate matter: results from the Children's Health Study. Environ Health Perspect, 2005. 113(11): p. 1638-44. 177. EPA, U., Basic Information about NO2. 2016, United States Environmental Protection Agency. 178. Amaducci, A. and J.W. Downs, Nitrogen Dioxide Toxicity. StatPearls [Internet], 2020. 179. McKay, C.A., Jr., Toxin-induced respiratory distress. Emerg Med Clin North Am, 2014. 32(1): p. 127-47. 180. Khreis, H., et al., Exposure to traffic-related air pollution and risk of development of childhood asthma: A systematic review and meta-analysis. Environ Int, 2017. 100: p. 1- 31. 165 181. Faustini, A., R. Rapp, and F. Forastiere, Nitrogen dioxide and mortality: review and meta-analysis of long-term studies. Eur Respir J, 2014. 44(3): p. 744-53. 182. Brauer, M., et al., A cohort study of traffic-related air pollution impacts on birth outcomes. Environ Health Perspect, 2008. 116(5): p. 680-6. 183. Evensen, E., et al., The relation between birthweight, childhood body mass index, and overweight and obesity in late adolescence: a longitudinal cohort study from Norway, The Tromsø Study, Fit Futures. BMJ open, 2017. 7(6). 184. EPA, Air Quality-National Summary. 2016. 185. Mohai, P., D. Pellow, and J.T. Roberts, Environmental Justice. Annual Review of Environment and Resources, 2009. 34(1): p. 405-430. 186. Zwickl, K., M. Ash, and J.K. Boyce, Regional variation in environmental inequality: Industrial air toxics exposure in US cities. Ecological Economics, 2014. 107: p. 494-509. 187. Cushing, L., et al., Racial/Ethnic Disparities in Cumulative Environmental Health Impacts in California: Evidence From a Statewide Environmental Justice Screening Tool (CalEnviroScreen 1.1). Am J Public Health, 2015. 105(11): p. 2341-8. 188. Weaver, G.M. and W.J. Gauderman, Traffic-related pollutants: Exposure and Health effects among Hispanic children. American journal of epidemiology, 2018. 187(1): p. 45- 52. 189. Fu, L., et al., The associations of air pollution exposure during pregnancy with fetal growth and anthropometric measurements at birth: a systematic review and meta- analysis. Environ Sci Pollut Res Int, 2019. 26(20): p. 20137-20147. 190. Iñiguez, C., et al., Prenatal exposure to traffic-related air pollution and fetal growth in a cohort of pregnant women. Occup Environ Med, 2012. 69(10): p. 736-44. 191. To, M. and S. Pereira, Routine Fetal Anomaly Scan. Twining's Textbook of Fetal Abnormalities E-Book, 2014. 192. Marsál, K., et al., Intrauterine growth curves based on ultrasonically estimated foetal weights. Acta Paediatr, 1996. 85(7): p. 843-8. 193. Lefler, J.S., et al., Air pollution and mortality in a large, representative US cohort: multiple-pollutant analyses, and spatial and temporal decompositions. Environmental Health, 2019. 18(1): p. 1-11. 194. Žaliūnas, B., et al., Fetal biometry: Relevance in obstetrical practice. Medicina, 2017. 53(6): p. 357-364. 166 195. Hadlock, F.P., et al., Estimation of fetal weight with the use of head, body, and femur measurements--a prospective study. Am J Obstet Gynecol, 1985. 151(3): p. 333-7. 196. Blake, B.E. and S.E. Fenton, Early life exposure to per-and polyfluoroalkyl substances (PFAS) and latent health outcomes: A review including the placenta as a target tissue and possible driver of peri-and postnatal effects. Toxicology, 2020: p. 152565. 197. Bové, H., et al., Ambient black carbon particles reach the fetal side of human placenta. Nature communications, 2019. 10(1): p. 1-7. 198. Perera, F. and J. Herbstman, Prenatal environmental exposures, epigenetics, and disease. Reproductive toxicology, 2011. 31(3): p. 363-373. 199. Kannan, S., et al., Exposures to airborne particulate matter and adverse perinatal outcomes: a biologically plausible mechanistic framework for exploring potential effect modification by nutrition. Environmental health perspectives, 2006. 114(11): p. 1636- 1642. 200. Saenen, N., et al., Air pollution-induced placental alterations: an interplay of oxidative stress, epigenetics, and the aging phenotype? Clinical epigenetics, 2019. 11(1): p. 1-14. 201. Maroziene, L. and R. Grazuleviciene, Maternal exposure to low-level air pollution and pregnancy outcomes: a population-based study. Environmental Health, 2002. 1(1): p. 1- 7. 202. Tabacova, S., D.D. Baird, and L. Balabaeva, Exposure to oxidized nitrogen: lipid peroxidation and neonatal health risk. Archives of Environmental Health: An International Journal, 1998. 53(3): p. 214-221. 203. Aguilera, I., et al., Prenatal exposure to traffic-related air pollution and ultrasound measures of fetal growth in the INMA Sabadell cohort. Environmental health perspectives, 2010. 118(5): p. 705-711. 204. Clemens, T., S. Turner, and C. Dibben, Maternal exposure to ambient air pollution and fetal growth in North-East Scotland: A population-based study using routine ultrasound scans. Environ Int, 2017. 107: p. 216-226. 205. Hansen, C.A., A.G. Barnett, and G. Pritchard, The effect of ambient air pollution during early pregnancy on fetal ultrasonic measurements during mid-pregnancy. Environ Health Perspect, 2008. 116(3): p. 362-9. 206. Iñiguez, C., et al., Prenatal exposure to traffic-related air pollution and fetal growth in a cohort of pregnant women. Occupational and environmental medicine, 2012. 69(10): p. 736-744. 167 207. Iñiguez, C., et al., Prenatal Exposure to NO2 and Ultrasound Measures of Fetal Growth in the Spanish INMA Cohort. Environ Health Perspect, 2016. 124(2): p. 235-42. 208. Lamichhane, D.K., et al., Air pollution exposure during pregnancy and ultrasound and birth measures of fetal growth: A prospective cohort study in Korea. Sci Total Environ, 2018. 619-620: p. 834-841. 209. Ritz, B., et al., Prenatal air pollution exposure and ultrasound measures of fetal growth in Los Angeles, California. Environ Res, 2014. 130: p. 7-13. 210. van den Hooven, E.H., et al., Air pollution exposure during pregnancy, ultrasound measures of fetal growth, and adverse birth outcomes: a prospective cohort study. Environ Health Perspect, 2012. 120(1): p. 150-6. 211. Wang, W., et al., Prenatal NO2 exposure and ultrasound measures of foetal growth: a prospective cohort study in Wuhan, China. Occupational and environmental medicine, 2017. 74(3): p. 204-210. 212. Iñiguez, C., et al., Prenatal exposure to NO2 and ultrasound measures of fetal growth in the Spanish INMA cohort. Environmental health perspectives, 2016. 124(2): p. 235-242. 213. Zhao, N., et al., Effects of prenatal exposure to ambient air pollutant PM10 on ultrasound-measured fetal growth. Int J Epidemiol, 2018. 47(4): p. 1072-1081. 214. Lin, L., et al., The associations of particulate matters with fetal growth in utero and birth weight: a birth cohort study in Beijing, China. Science of the Total Environment, 2020. 709: p. 136246. 215. Cao, Z., et al., Maternal exposure to ambient fine particulate matter and fetal growth in Shanghai, China. Environmental health, 2019. 18(1): p. 1-8. 216. Leung, M., et al., Exposure to PM 2.5 during Pregnancy and Fetal Growth in Eastern Massachusetts, USA. Environmental health perspectives, 2022. 130(1): p. 017004. 217. Mohai, P. and R. Saha, Racial inequality in the distribution of hazardous waste: A national-level reassessment. Social problems, 2007. 54(3): p. 343-370. 218. Martin, J.A., et al., Births: Final Data for 2019. Natl Vital Stat Rep, 2021. 70(2): p. 1-51. 219. Bastain, T.M., et al., Study Design, Protocol and Profile of the Maternal And Developmental Risks from Environmental and Social Stressors (MADRES) Pregnancy Cohort: a Prospective Cohort Study in Predominantly Low-Income Hispanic Women in Urban Los Angeles. BMC Pregnancy Childbirth, 2019. 19(1): p. 189. 220. Obstetricians, A.C.o. and Gynecologists, Committee opinion no 700: methods for estimating the due date. Obstet Gynecol, 2017. 129(5): p. e150-e154. 168 221. Aris, I.M., et al., A 2017 US reference for singleton birth weight percentiles using obstetric estimates of gestation. Pediatrics, 2019. 144(1). 222. Salomon, L., et al., ISUOG Practice Guidelines: ultrasound assessment of fetal biometry and growth. Ultrasound in obstetrics & gynecology, 2019. 53(6): p. 715-723. 223. Hadlock, F.P., et al., Estimation of fetal weight with the use of head, body, and femur measurements—a prospective study. American journal of obstetrics and gynecology, 1985. 151(3): p. 333-337. 224. Houde, M.; De Silva, A. O.; Muir, D. C.; Letcher, R. J., Monitoring of perfluorinated compounds in aquatic biota: an updated review. Environmental science & technology 2011, 45, (19), 7962-73. 225. Sinclair, G. M.; Long, S. M.; Jones, O. A., What are the effects of PFAS exposure at environmentally relevant concentrations? Chemosphere 2020, 127340. 226. Bassler, J.; Ducatman, A.; Elliott, M.; Wen, S.; Wahlang, B.; Barnett, J.; Cave, M. C., Environmental perfluoroalkyl acid exposures are associated with liver disease characterized by apoptosis and altered serum adipocytokines. Environmental Pollution 2019, 247, 1055-1063. 227. Chen, J.; Zheng, L.; Tian, L.; Wang, N.; Lei, L.; Wang, Y.; Dong, Q.; Huang, C.; Yang, D., Chronic PFOS exposure disrupts thyroid structure and function in zebrafish. Bulletin of environmental contamination and toxicology 2018, 101, (1), 75-79. 228. Chou, H.-C.; Wen, L.-L.; Chang, C.-C.; Lin, C.-Y.; Jin, L.; Juan, S.-H., From the cover: l-carnitine via PPARγ-and Sirt1-dependent mechanisms attenuates epithelial- mesenchymal transition and renal fibrosis caused by perfluorooctanesulfonate. Toxicological Sciences 2017, 160, (2), 217-229. 229. Zeng, Z.; Song, B.; Xiao, R.; Zeng, G.; Gong, J.; Chen, M.; Xu, P.; Zhang, P.; Shen, M.; Yi, H., Assessing the human health risks of perfluorooctane sulfonate by in vivo and in vitro studies. Environment international 2019, 126, 598-610. 230. Koustas, E.; Lam, J.; Sutton, P.; Johnson, P. I.; Atchley, D. S.; Sen, S.; Robinson, K. A.; Axelrad, D. A.; Woodruff, T. J., The Navigation Guide - evidence-based medicine meets environmental health: systematic review of nonhuman evidence for PFOA effects on fetal growth. Environmental health perspectives 2014, 122, (10), 1015-27. 231. Negri, E.; Metruccio, F.; Guercio, V.; Tosti, L.; Benfenati, E.; Bonzi, R.; La Vecchia, C.; Moretto, A., Exposure to PFOA and PFOS and fetal growth: a critical merging of toxicological and epidemiological data. Critical reviews in toxicology 2017, 47, (6), 489- 515. 169 232. EPA, U., 2010/2015 PFOA Stewardship Program. In United States Environmental Protection Agency Washington, DC: 2006. 233. Olsen, G. W.; Burris, J. M.; Ehresman, D. J.; Froehlich, J. W.; Seacat, A. M.; Butenhoff, J. L.; Zobel, L. R., Half-life of serum elimination of perfluorooctanesulfonate,perfluorohexanesulfonate, and perfluorooctanoate in retired fluorochemical production workers. Environmental health perspectives 2007, 115, (9), 1298-305. 234. Calafat, A. M.; Wong, L. Y.; Kuklenyik, Z.; Reidy, J. A.; Needham, L. L., Polyfluoroalkyl chemicals in the U.S. population: data from the National Health and Nutrition Examination Survey (NHANES) 2003-2004 and comparisons with NHANES 1999-2000. Environmental health perspectives 2007, 115, (11), 1596-602. 235. Ye, X.; Kato, K.; Wong, L.-Y.; Jia, T.; Kalathil, A.; Latremouille, J.; Calafat, A. M., Per- and polyfluoroalkyl substances in sera from children 3 to 11 years of age participating in the National Health and Nutrition Examination Survey 2013–2014. International journal of hygiene and environmental health 2018, 221, (1), 9-16. 236. Fair, P. A.; Wolf, B.; White, N. D.; Arnott, S. A.; Kannan, K.; Karthikraj, R.; Vena, J. E., Perfluoroalkyl substances (PFASs) in edible fish species from Charleston Harbor and tributaries, South Carolina, United States: Exposure and risk assessment. Environmental research 2019, 171, 266-277. 237. Hoffman, K.; Webster, T. F.; Bartell, S. M.; Weisskopf, M. G.; Fletcher, T.; Vieira, V. M., Private drinking water wells as a source of exposure to perfluorooctanoic acid (PFOA) in communities surrounding a fluoropolymer production facility. Environmental health perspectives 2011, 119, (1), 92-7. 238. Schaider, L. A.; Balan, S. A.; Blum, A.; Andrews, D. Q.; Strynar, M. J.; Dickinson, M. E.; Lunderberg, D. M.; Lang, J. R.; Peaslee, G. F., Fluorinated Compounds in U.S. Fast Food Packaging. Environmental science & technology letters 2017, 4, (3), 105-111. 239. Sunderland, E. M.; Hu, X. C.; Dassuncao, C.; Tokranov, A. K.; Wagner, C. C.; Allen, J. G., A review of the pathways of human exposure to poly-and perfluoroalkyl substances (PFASs) and present understanding of health effects. Journal of exposure science & environmental epidemiology 2019, 29, (2), 131-147. 240. Tanner, E. M.; Bloom, M. S.; Wu, Q.; Kannan, K.; Yucel, R. M.; Shrestha, S.; Fitzgerald, E. F., Occupational exposure to perfluoroalkyl substances and serum levels of perfluorooctanesulfonic acid (PFOS) and perfluorooctanoic acid (PFOA) in an aging population from upstate New York: a retrospective cohort study. International archives of occupational and environmental health 2018, 91, (2), 145-154. 241. Tittlemier, S. A.; Pepper, K.; Seymour, C.; Moisey, J.; Bronson, R.; Cao, X. L.; Dabeka, R. W., Dietary exposure of Canadians to perfluorinated carboxylates and perfluorooctane 170 sulfonate via consumption of meat, fish, fast foods, and food items prepared in their packaging. Journal of agricultural and food chemistry 2007, 55, (8), 3203-10. 242. Mamsen, L. S.; Jönsson, B. A. G.; Lindh, C. H.; Olesen, R. H.; Larsen, A.; Ernst, E.; Kelsey, T. W.; Andersen, C. Y., Concentration of perfluorinated compounds and cotinine in human foetal organs, placenta, and maternal plasma. The Science of the total environment 2017, 596-597, 97-105. 243. Blake, B. E.; Fenton, S. E., Early life exposure to per- and polyfluoroalkyl substances (PFAS) and latent health outcomes: A review including the placenta as a target tissue and possible driver of peri- and postnatal effects. Toxicology 2020, 443, 152565. 244. Szilagyi, J. T.; Avula, V.; Fry, R. C., Perfluoroalkyl Substances (PFAS) and Their Effects on the Placenta, Pregnancy, and Child Development: a Potential Mechanistic Role for Placental Peroxisome Proliferator-Activated Receptors (PPARs). Current environmental health reports 2020, 7, (3), 222-230. 245. Du, G.; Hu, J.; Huang, H.; Qin, Y.; Han, X.; Wu, D.; Song, L.; Xia, Y.; Wang, X., Perfluorooctane sulfonate (PFOS) affects hormone receptor activity, steroidogenesis, and expression of endocrine-related genes in vitro and in vivo. Environmental toxicology and chemistry 2013, 32, (2), 353-60. 246. Kaijser, M.; Granath, F.; Jacobsen, G.; Cnattingius, S.; Ekbom, A., Maternal pregnancy estriol levels in relation to anamnestic and fetal anthropometric data. Epidemiology (Cambridge, Mass.) 2000, 11, (3), 315-9. 247. Benninghoff, A. D.; Bisson, W. H.; Koch, D. C.; Ehresman, D. J.; Kolluri, S. K.; Williams, D. E., Estrogen-like activity of perfluoroalkyl acids in vivo and interaction with human and rainbow trout estrogen receptors in vitro. Toxicological sciences : an official journal of the Society of Toxicology 2011, 120, (1), 42-58. 248. Henry, N. D.; Fair, P. A., Comparison of in vitro cytotoxicity, estrogenicity and anti- estrogenicity of triclosan, perfluorooctane sulfonate and perfluorooctanoic acid. Journal of applied toxicology : JAT 2013, 33, (4), 265-72. 249. Kjeldsen, L. S.; Bonefeld-Jørgensen, E. C., Perfluorinated compounds affect the function of sex hormone receptors. Environmental science and pollution research international 2013, 20, (11), 8031-44. 250. Suh, C. H.; Cho, N. K.; Lee, C. K.; Lee, C. H.; Kim, D. H.; Kim, J. H.; Son, B. C.; Lee, J. T., Perfluorooctanoic acid-induced inhibition of placental prolactin-family hormone and fetal growth retardation in mice. Molecular and cellular endocrinology 2011, 337, (1-2), 7-15. 251. Glazier, J. D.; Cetin, I.; Perugino, G.; Ronzoni, S.; Grey, A. M.; Mahendran, D.; Marconi, A. M.; Pardi, G.; Sibley, C. P., Association between the activity of the system A 171 amino acid transporter in the microvillous plasma membrane of the human placenta and severity of fetal compromise in intrauterine growth restriction. Pediatr Res 1997, 42, (4), 514-9. 252. Koskela, A.; Koponen, J.; Lehenkari, P.; Viluksela, M.; Korkalainen, M.; Tuukkanen, J., Perfluoroalkyl substances in human bone: concentrations in bones and effects on bone cell differentiation. Scientific reports 2017, 7, (1), 6841. 253. Bach, C. C.; Bech, B. H.; Brix, N.; Nohr, E. A.; Bonde, J. P.; Henriksen, T. B., Perfluoroalkyl and polyfluoroalkyl substances and human fetal growth: a systematic review. Critical reviews in toxicology 2015, 45, (1), 53-67. 254. Johnson, P. I.; Sutton, P.; Atchley, D. S.; Koustas, E.; Lam, J.; Sen, S.; Robinson, K. A.; Axelrad, D. A.; Woodruff, T. J., The Navigation Guide - evidence-based medicine meets environmental health: systematic review of human evidence for PFOA effects on fetal growth. Environmental health perspectives 2014, 122, (10), 1028-39. 255. Steenland, K.; Barry, V.; Savitz, D., Serum Perfluorooctanoic Acid and Birthweight: An Updated Meta-analysis With Bias Analysis. Epidemiology (Cambridge, Mass.) 2018, 29, (6), 765-776. 256. Dzierlenga, M. W.; Crawford, L.; Longnecker, M. P., Birth weight and perfluorooctane sulfonic acid: a random-effects meta-regression analysis. In LWW: 2020. 257. Chen, L.; Tong, C.; Huo, X.; Zhang, J.; Tian, Y.; Cohort, S. B., Prenatal exposure to perfluoroalkyl and polyfluoroalkyl substances and birth outcomes: A longitudinal cohort with repeated measurements. Chemosphere 2021, 267, 128899. 258. Maisonet, M.; Terrell, M. L.; McGeehin, M. A.; Christensen, K. Y.; Holmes, A.; Calafat, A. M.; Marcus, M., Maternal concentrations of polyfluoroalkyl compounds during pregnancy and fetal and postnatal growth in British girls. Environmental health perspectives 2012, 120, (10), 1432-7. 259. Wikström, S.; Lin, P. I.; Lindh, C. H.; Shu, H.; Bornehag, C. G., Maternal serum levels of perfluoroalkyl substances in early pregnancy and offspring birth weight. Pediatr Res 2020, 87, (6), 1093-1099. 260. Meng, Q.; Inoue, K.; Ritz, B.; Olsen, J.; Liew, Z., Prenatal exposure to perfluoroalkyl substances and birth outcomes; an updated analysis from the Danish National Birth Cohort. International journal of environmental research and public health 2018, 15, (9), 1832. 261. Marks, K. J.; Cutler, A. J.; Jeddy, Z.; Northstone, K.; Kato, K.; Hartman, T. J., Maternal serum concentrations of perfluoroalkyl substances and birth size in British boys. International journal of hygiene and environmental health 2019, 222, (5), 889-895. 172 262. Kashino, I.; Sasaki, S.; Okada, E.; Matsuura, H.; Goudarzi, H.; Miyashita, C.; Okada, E.; Ito, Y. M.; Araki, A.; Kishi, R., Prenatal exposure to 11 perfluoroalkyl substances and fetal growth: A large-scale, prospective birth cohort study. Environment international 2020, 136, 105355. 263. Kashino, I.; Sasaki, S.; Okada, E.; Matsuura, H.; Goudarzi, H.; Miyashita, C.; Okada, E.; Ito, Y. M.; Araki, A.; Kishi, R., Prenatal exposure to 11 perfluoroalkyl substances and fetal growth: A large-scale, prospective birth cohort study. Environ Int 2020, 136, 105355. 264. Meng, Q.; Inoue, K.; Ritz, B.; Olsen, J.; Liew, Z., Prenatal Exposure to Perfluoroalkyl Substances and Birth Outcomes; An Updated Analysis from the Danish National Birth Cohort. International journal of environmental research and public health 2018, 15, (9). 265. Manzano-Salgado, C. B.; Casas, M.; Lopez-Espinosa, M. J.; Ballester, F.; Iñiguez, C.; Martinez, D.; Costa, O.; Santa-Marina, L.; Pereda-Pereda, E.; Schettgen, T.; Sunyer, J.; Vrijheid, M., Prenatal exposure to perfluoroalkyl substances and birth outcomes in a Spanish birth cohort. Environ Int 2017, 108, 278-284. 266. Barker, D. J.; Hales, C. N.; Fall, C. H.; Osmond, C.; Phipps, K.; Clark, P. M., Type 2 (non-insulin-dependent) diabetes mellitus, hypertension and hyperlipidaemia (syndrome X): relation to reduced fetal growth. Diabetologia 1993, 36, (1), 62-7. 267. Evensen, E.; Emaus, N.; Kokkvoll, A.; Wilsgaard, T.; Furberg, A.-S.; Skeie, G., The relation between birthweight, childhood body mass index, and overweight and obesity in late adolescence: a longitudinal cohort study from Norway, The Tromsø Study, Fit Futures. BMJ open 2017, 7, (6). 268. Jornayvaz, F. R.; Vollenweider, P.; Bochud, M.; Mooser, V.; Waeber, G.; Marques- Vidal, P., Low birth weight leads to obesity, diabetes and increased leptin levels in adults: the CoLaus study. Cardiovascular diabetology 2016, 15, (1), 1-10. 269. Martin, J. A.; Hamilton, B. E.; Osterman, M. J. K.; Driscoll, A. K., Births: Final Data for 2019. National vital statistics reports : from the Centers for Disease Control and Prevention, National Center for Health Statistics, National Vital Statistics System 2021, 70, (2), 1-51. 270. Bastain, T. M.; Chavez, T.; Habre, R.; Girguis, M. S.; Grubbs, B.; Toledo-Corral, C.; Amadeus, M.; Farzan, S. F.; Al-Marayati, L.; Lerner, D.; Noya, D.; Quimby, A.; Twogood, S.; Wilson, M.; Chatzi, L.; Cousineau, M.; Berhane, K.; Eckel, S. P.; Lurmann, F.; Johnston, J.; Dunton, G. F.; Gilliland, F.; Breton, C., Study Design, Protocol and Profile of the Maternal And Developmental Risks from Environmental and Social Stressors (MADRES) Pregnancy Cohort: a Prospective Cohort Study in Predominantly Low-Income Hispanic Women in Urban Los Angeles. BMC pregnancy and childbirth 2019, 19, (1), 189. 173 271. Honda, M.; Robinson, M.; Kannan, K., A rapid method for the analysis of perfluorinated alkyl substances in serum by hybrid solid-phase extraction. Environmental Chemistry 2018, 15, (2), 92-99. 272. Practice, C. o. O.; Medicine, A. I. o. U. i.; Medicine, S. f. M.-F., Committee opinion no 700: methods for estimating the due date. Obstetrics and gynecology 2017, 129, (5), e150-e154. 273. Aris, I. M.; Kleinman, K. P.; Belfort, M. B.; Kaimal, A.; Oken, E., A 2017 US Reference for Singleton Birth Weight Percentiles Using Obstetric Estimates of Gestation. Pediatrics 2019, 144, (1). 274. Textor, J.; van der Zander, B.; Gilthorpe, M. S.; Liskiewicz, M.; Ellison, G. T., Robust causal inference using directed acyclic graphs: the R package 'dagitty'. International journal of epidemiology 2016, 45, (6), 1887-1894. 275. Hornung, R. W.; Reed, L. D., Estimation of average concentration in the presence of nondetectable values. Applied occupational and environmental hygiene 1990, 5, (1), 46- 51. 276. Wilcox, A. J.; Weinberg, C. R.; Basso, O., On the pitfalls of adjusting for gestational age at birth. American journal of epidemiology 2011, 174, (9), 1062-8. 277. Eick, S. M.; Hom Thepaksorn, E. K.; Izano, M. A.; Cushing, L. J.; Wang, Y.; Smith, S. C.; Gao, S.; Park, J. S.; Padula, A. M.; DeMicco, E.; Valeri, L.; Woodruff, T. J.; Morello- Frosch, R., Associations between prenatal maternal exposure to per- and polyfluoroalkyl substances (PFAS) and polybrominated diphenyl ethers (PBDEs) and birth outcomes among pregnant women in San Francisco. Environmental health : a global access science source 2020, 19, (1), 100. 278. Sagiv, S. K.; Rifas-Shiman, S. L.; Webster, T. F.; Mora, A. M.; Harris, M. H.; Calafat, A. M.; Ye, X.; Gillman, M. W.; Oken, E., Sociodemographic and Perinatal Predictors of Early Pregnancy Per- and Polyfluoroalkyl Substance (PFAS) Concentrations. Environmental science & technology 2015, 49, (19), 11849-58. 279. Starling, A. P.; Adgate, J. L.; Hamman, R. F.; Kechris, K.; Calafat, A. M.; Ye, X.; Dabelea, D., Perfluoroalkyl Substances during Pregnancy and Offspring Weight and Adiposity at Birth: Examining Mediation by Maternal Fasting Glucose in the Healthy Start Study. Environmental health perspectives 2017, 125, (6), 067016. 280. Fenton, S. E.; Ducatman, A.; Boobis, A.; DeWitt, J. C.; Lau, C.; Ng, C.; Smith, J. S.; Roberts, S. M., Per- and Polyfluoroalkyl Substance Toxicity and Human Health Review: Current State of Knowledge and Strategies for Informing Future Research. Environmental toxicology and chemistry 2021, 40, (3), 606-630. 174 281. Barbarossa, A.; Masetti, R.; Gazzotti, T.; Zama, D.; Astolfi, A.; Veyrand, B.; Pession, A.; Pagliuca, G., Perfluoroalkyl substances in human milk: a first survey in Italy. Environment international 2013, 51, 27-30. 282. Kato, K.; Wong, L. Y.; Chen, A.; Dunbar, C.; Webster, G. M.; Lanphear, B. P.; Calafat, A. M., Changes in serum concentrations of maternal poly- and perfluoroalkyl substances over the course of pregnancy and predictors of exposure in a multiethnic cohort of Cincinnati, Ohio pregnant women during 2003-2006. Environmental science & technology 2014, 48, (16), 9600-8. 283. Mondal, D.; Weldon, R. H.; Armstrong, B. G.; Gibson, L. J.; Lopez-Espinosa, M. J.; Shin, H. M.; Fletcher, T., Breastfeeding: a potential excretion route for mothers and implications for infant exposure to perfluoroalkyl acids. Environmental health perspectives 2014, 122, (2), 187-92. 284. Tao, L.; Kannan, K.; Wong, C. M.; Arcaro, K. F.; Butenhoff, J. L., Perfluorinated compounds in human milk from Massachusetts, U.S.A. Environmental science & technology 2008, 42, (8), 3096-101. 285. Thomsen, C.; Haug, L. S.; Stigum, H.; Frøshaug, M.; Broadwell, S. L.; Becher, G., Changes in concentrations of perfluorinated compounds, polybrominated diphenyl ethers, and polychlorinated biphenyls in Norwegian breast-milk during twelve months of lactation. Environmental science & technology 2010, 44, (24), 9550-6. 286. Winkens, K.; Vestergren, R.; Berger, U.; Cousins, I. T., Early life exposure to per-and polyfluoroalkyl substances (PFASs): A critical review. Emerging Contaminants 2017, 3, (2), 55-68. 287. Boronow, K. E.; Brody, J. G.; Schaider, L. A.; Peaslee, G. F.; Havas, L.; Cohn, B. A., Serum concentrations of PFASs and exposure-related behaviors in African American and non-Hispanic white women. Journal of exposure science & environmental epidemiology 2019, 29, (2), 206-217. 288. Park, S. K.; Peng, Q.; Ding, N.; Mukherjee, B.; Harlow, S. D., Determinants of per- and polyfluoroalkyl substances (PFAS) in midlife women: Evidence of racial/ethnic and geographic differences in PFAS exposure. Environmental research 2019, 175, 186-199. 289. Cushing, L.; Faust, J.; August, L. M.; Cendak, R.; Wieland, W.; Alexeeff, G., Racial/Ethnic Disparities in Cumulative Environmental Health Impacts in California: Evidence From a Statewide Environmental Justice Screening Tool (CalEnviroScreen 1.1). American journal of public health 2015, 105, (11), 2341-8. 290. Brantsæter, A. L.; Whitworth, K. W.; Ydersbond, T. A.; Haug, L. S.; Haugen, M.; Knutsen, H. K.; Thomsen, C.; Meltzer, H. M.; Becher, G.; Sabaredzovic, A.; Hoppin, J. A.; Eggesbø, M.; Longnecker, M. P., Determinants of plasma concentrations of perfluoroalkyl substances in pregnant Norwegian women. Environ Int 2013, 54, 74-84. 175 291. Tsai, M. S.; Miyashita, C.; Araki, A.; Itoh, S.; Bamai, Y. A.; Goudarzi, H.; Okada, E.; Kashino, I.; Matsuura, H.; Kishi, R., Determinants and Temporal Trends of Perfluoroalkyl Substances in Pregnant Women: The Hokkaido Study on Environment and Children's Health. International journal of environmental research and public health 2018, 15, (5). 292. Darrow, L. A.; Stein, C. R.; Steenland, K., Serum perfluorooctanoic acid and perfluorooctane sulfonate concentrations in relation to birth outcomes in the Mid-Ohio Valley, 2005-2010. Environmental health perspectives 2013, 121, (10), 1207-13. 293. Fei, C.; McLaughlin, J. K.; Tarone, R. E.; Olsen, J., Perfluorinated chemicals and fetal growth: a study within the Danish National Birth Cohort. Environmental health perspectives 2007, 115, (11), 1677-1682. 294. Hamm, M. P.; Cherry, N. M.; Chan, E.; Martin, J. W.; Burstyn, I., Maternal exposure to perfluorinated acids and fetal growth. J Expo Sci Environ Epidemiol 2010, 20, (7), 589- 97. 295. Monroy, R.; Morrison, K.; Teo, K.; Atkinson, S.; Kubwabo, C.; Stewart, B.; Foster, W. G., Serum levels of perfluoroalkyl compounds in human maternal and umbilical cord blood samples. Environmental research 2008, 108, (1), 56-62. 296. Washino, N.; Saijo, Y.; Sasaki, S.; Kato, S.; Ban, S.; Konishi, K.; Ito, R.; Nakata, A.; Iwasaki, Y.; Saito, K.; Nakazawa, H.; Kishi, R., Correlations between prenatal exposure to perfluorinated chemicals and reduced fetal growth. Environmental health perspectives 2009, 117, (4), 660-7. 297. Chen, M. H.; Ha, E. H.; Wen, T. W.; Su, Y. N.; Lien, G. W.; Chen, C. Y.; Chen, P. C.; Hsieh, W. S., Perfluorinated compounds in umbilical cord blood and adverse birth outcomes. PloS one 2012, 7, (8), e42474. 298. Calafat, A.; Kato, K.; Ye, X. In Urine and Serum Biomarkers of Per-and Polyfluoroalkyl Substances (PFAS) and Fluorinated Alternatives for Human Exposure Assessment, ISEE Conference Abstracts, 2018; 2018. 299. Ehresman, D. J.; Froehlich, J. W.; Olsen, G. W.; Chang, S. C.; Butenhoff, J. L., Comparison of human whole blood, plasma, and serum matrices for the determination of perfluorooctanesulfonate (PFOS), perfluorooctanoate (PFOA), and other fluorochemicals. Environmental research 2007, 103, (2), 176-84. 300. Verner, M. A.; Loccisano, A. E.; Morken, N. H.; Yoon, M.; Wu, H.; McDougall, R.; Maisonet, M.; Marcus, M.; Kishi, R.; Miyashita, C.; Chen, M. H.; Hsieh, W. S.; Andersen, M. E.; Clewell, H. J., 3rd; Longnecker, M. P., Associations of Perfluoroalkyl Substances (PFAS) with Lower Birth Weight: An Evaluation of Potential Confounding 176 by Glomerular Filtration Rate Using a Physiologically Based Pharmacokinetic Model (PBPK). Environmental health perspectives 2015, 123, (12), 1317-24. 301. Salas, S. a. P.; Marshall, G.; Gutié rrez, B. L.; Rosso, P., Time course of maternal plasma volume and hormonal changes in women with preeclampsia or fetal growth restriction. Hypertension 2006, 47, (2), 203-208. 302. Sinclair, G.M., S.M. Long, and O.A.H. Jones, What are the effects of PFAS exposure at environmentally relevant concentrations? Chemosphere, 2020. 258: p. 127340. 303. Dauchy, X., et al., Per- and polyfluoroalkyl substances in firefighting foam concentrates and water samples collected near sites impacted by the use of these foams. Chemosphere, 2017. 183: p. 53-61. 304. Schaider, L.A., et al., Fluorinated compounds in US fast food packaging. Environmental science & technology letters, 2017. 4(3): p. 105-111. 305. Ding, G. and W.J. Peijnenburg, Physicochemical properties and aquatic toxicity of poly- and perfluorinated compounds. Critical reviews in environmental science and technology, 2013. 43(6): p. 598-678. 306. De Silva, A.O., et al., PFAS exposure pathways for humans and wildlife: a synthesis of current knowledge and key gaps in understanding. Environmental Toxicology and Chemistry, 2021. 40(3): p. 631-657. 307. Bassler, J., et al., Environmental perfluoroalkyl acid exposures are associated with liver disease characterized by apoptosis and altered serum adipocytokines. Environmental Pollution, 2019. 247: p. 1055-1063. 308. Chen, J., et al., Chronic PFOS exposure disrupts thyroid structure and function in zebrafish. Bulletin of environmental contamination and toxicology, 2018. 101(1): p. 75- 79. 309. Chou, H.-C., et al., From the cover: l-carnitine via PPARγ-and Sirt1-dependent mechanisms attenuates epithelial-mesenchymal transition and renal fibrosis caused by perfluorooctanesulfonate. Toxicological Sciences, 2017. 160(2): p. 217-229. 310. Zeng, Z., et al., Assessing the human health risks of perfluorooctane sulfonate by in vivo and in vitro studies. Environment international, 2019. 126: p. 598-610. 311. EPA, 2015 PFOA stewardship program. US Environmental Protection Agency, 2010. 312. Olsen, G.W., et al., Half-life of serum elimination of perfluorooctanesulfonate, perfluorohexanesulfonate, and perfluorooctanoate in retired fluorochemical production workers. Environmental health perspectives, 2007. 115(9): p. 1298-1305. 177 313. Calafat, A.M., et al., Polyfluoroalkyl chemicals in the US population: data from the National Health and Nutrition Examination Survey (NHANES) 2003–2004 and comparisons with NHANES 1999–2000. Environmental health perspectives, 2007. 115(11): p. 1596-1602. 314. Ye, X., et al., Per-and polyfluoroalkyl substances in sera from children 3 to 11 years of age participating in the National Health and Nutrition Examination Survey 2013–2014. International journal of hygiene and environmental health, 2018. 221(1): p. 9-16. 315. Varshavsky, J., et al., Heightened susceptibility: A review of how pregnancy and chemical exposures influence maternal health. Reproductive toxicology, 2020. 92: p. 14- 56. 316. Blake, B.E. and S.E. Fenton, Early life exposure to per- and polyfluoroalkyl substances (PFAS) and latent health outcomes: A review including the placenta as a target tissue and possible driver of peri- and postnatal effects. Toxicology, 2020. 443: p. 152565. 317. Mamsen, L.S., et al., Concentrations of perfluoroalkyl substances (PFASs) in human embryonic and fetal organs from first, second, and third trimester pregnancies. Environment international, 2019. 124: p. 482-492. 318. Barker, D.J., et al., Type 2 (non-insulin-dependent) diabetes mellitus, hypertension and hyperlipidaemia (syndrome X): relation to reduced fetal growth. Diabetologia, 1993. 36(1): p. 62-67. 319. Barker, D.J., et al., Trajectories of growth among children who have coronary events as adults. N Engl J Med, 2005. 353(17): p. 1802-9. 320. Gluckman, P.D., et al., Effect of in utero and early-life conditions on adult health and disease. N Engl J Med, 2008. 359(1): p. 61-73. 321. Bartels, H.C., et al., Fetal growth trajectories and their association with maternal and child characteristics. J Matern Fetal Neonatal Med, 2020. 33(14): p. 2427-2433. 322. Rückinger, S., et al., Growth in utero and body mass index at age 5 years in children of smoking and non-smoking mothers. Early Hum Dev, 2010. 86(12): p. 773-7. 323. Von Ehrenstein, O.S., R.T. Mikolajczyk, and J. Zhang, Timing and trajectories of fetal growth related to cognitive development in childhood. American Journal of Epidemiology, 2009. 170(11): p. 1388-1395. 324. Bach, C.C., et al., Perfluoroalkyl and polyfluoroalkyl substances and human fetal growth: a systematic review. Crit Rev Toxicol, 2015. 45(1): p. 53-67. 325. Dzierlenga, M.W., L. Crawford, and M.P. Longnecker, Birth weight and perfluorooctane sulfonic acid: a random-effects meta-regression analysis. 2020, LWW. 178 326. Negri, E., et al., Exposure to PFOA and PFOS and fetal growth: a critical merging of toxicological and epidemiological data. Critical reviews in toxicology, 2017. 47(6): p. 489-515. 327. Steenland, K., V. Barry, and D. Savitz, Serum Perfluorooctanoic Acid and Birthweight: An Updated Meta-analysis With Bias Analysis. Epidemiology, 2018. 29(6): p. 765-776. 328. Costa, O., et al., First-trimester maternal concentrations of polyfluoroalkyl substances and fetal growth throughout pregnancy. Environment international, 2019. 130: p. 104830. 329. Ouidir, M., et al., Association of maternal exposure to persistent organic pollutants in early pregnancy with fetal growth. JAMA pediatrics, 2020. 174(2): p. 149-161. 330. Osterman, M., et al., Births: Final Data for 2020. Natl Vital Stat Rep, 2021. 70(17): p. 1- 50. 331. Williams, D.R., Stress and the Mental Health of Populations of Color: Advancing Our Understanding of Race-related Stressors. J Health Soc Behav, 2018. 59(4): p. 466-485. 332. Lima, S.A.M., et al., Is the risk of low birth weight or preterm labor greater when maternal stress is experienced during pregnancy? A systematic review and meta-analysis of cohort studies. PLoS One, 2018. 13(7): p. e0200594. 333. Staneva, A., et al., The effects of maternal depression, anxiety, and perceived stress during pregnancy on preterm birth: A systematic review. Women Birth, 2015. 28(3): p. 179-93. 334. Vesterinen, H.M., et al., Cumulative effects of prenatal-exposure to exogenous chemicals and psychosocial stress on fetal growth: Systematic-review of the human and animal evidence. PLoS One, 2017. 12(7): p. e0176331. 335. Bastain, T.M., et al., Study design, protocol and profile of the Maternal and Developmental Risks from Environmental and Social Stressors (MADRES) pregnancy cohort: a prospective cohort study in predominantly low-income Hispanic women in urban Los Angeles. BMC pregnancy and childbirth, 2019. 19(1): p. 1-16. 336. Honda, M., M. Robinson, and K. Kannan, A rapid method for the analysis of perfluorinated alkyl substances in serum by hybrid solid-phase extraction. Environmental Chemistry, 2018. 15(2): p. 92-99. 337. Hadlock, F.P., et al., Estimation of fetal weight with the use of head, body, and femur measurements--a prospective study. Am J Obstet Gynecol, 1985. 151(3): p. 333-7. 338. Cohen, S., Perceived stress in a probability sample of the United States. 1988. 179 339. Committee on Obstetric Practice, t.A.I.o.U.i.M. and t.S.f.M.-F. Medicine, Committee opinion No 700: Methods for estimating the due date. Obstet Gynecol, 2017. 129(5): p. e150-e154. 340. Textor, J., et al., Robust causal inference using directed acyclic graphs: the R package ‘dagitty’. International journal of epidemiology, 2016. 45(6): p. 1887-1894. 341. Pinheiro, J., et al., The Nlme package: linear and nonlinear mixed effects models, R version 3. R package version, 2012. 6. 342. Woodruff, T.J., A.R. Zota, and J.M. Schwartz, Environmental chemicals in pregnant women in the United States: NHANES 2003–2004. Environmental health perspectives, 2011. 119(6): p. 878-885. 343. Gale, C.R., et al., The influence of head growth in fetal life, infancy, and childhood on intelligence at the ages of 4 and 8 years. Pediatrics, 2006. 118(4): p. 1486-92. 344. Villar, J., et al., Fetal cranial growth trajectories are associated with growth and neurodevelopment at 2 years of age: INTERBIO-21st Fetal Study. Nat Med, 2021. 27(4): p. 647-652. 345. Zheng, T., et al., Effects of Environmental Exposures on Fetal and Childhood Growth Trajectories. Ann Glob Health, 2016. 82(1): p. 41-99. 346. Apelberg, B.J., et al., Cord serum concentrations of perfluorooctane sulfonate (PFOS) and perfluorooctanoate (PFOA) in relation to weight and size at birth. Environ Health Perspect, 2007. 115(11): p. 1670-6. 347. Xiao, C., et al., Associations of Exposure to Perfluoroalkyl Substances With Thyroid Hormone Concentrations and Birth Size. J Clin Endocrinol Metab, 2020. 105(3): p. 735- 45. 348. Eick, S.M., et al., Prenatal PFAS and psychosocial stress exposures in relation to fetal growth in two pregnancy cohorts: applying environmental mixture methods to chemical and non-chemical stressors. Environment International, 2022: p. 107238. 349. Dreyer, A.F., et al., Perfluoroalkyl substance exposure early in pregnancy was negatively associated with late pregnancy cortisone levels. The Journal of Clinical Endocrinology & Metabolism, 2020. 105(8): p. e2834-e2844. 350. Eick, S.M., et al., Joint effects of prenatal exposure to per-and poly-fluoroalkyl substances and psychosocial stressors on corticotropin-releasing hormone during pregnancy. Journal of exposure science & environmental epidemiology, 2022. 32(1): p. 27-36. 180 351. Nguyen, V.K., et al., A comprehensive analysis of racial disparities in chemical biomarker concentrations in United States women, 1999-2014. Environ Int, 2020. 137: p. 105496. 352. Fei, C., et al., Perfluorinated chemicals and fetal growth: a study within the Danish National Birth Cohort. Environ Health Perspect, 2007. 115(11): p. 1677-82. 353. Chen, L., et al., Prenatal exposure to perfluoroalkyl and polyfluoroalkyl substances and birth outcomes: A longitudinal cohort with repeated measurements. Chemosphere, 2021. 267: p. 128899. 354. Schraufnagel, D.E., et al., Air pollution and noncommunicable diseases: A review by the Forum of International Respiratory Societies’ Environmental Committee, Part 2: Air pollution and organ systems. Chest, 2019. 155(2): p. 417-426. 355. State of the Air. 2021, American Lung Association. 356. Dedoussi, I.C., et al., Premature mortality related to United States cross-state air pollution. Nature, 2020. 578(7794): p. 261-265. 357. Hayes, R.B., et al., PM2.5 air pollution and cause-specific cardiovascular disease mortality. Int J Epidemiol, 2020. 49(1): p. 25-35. 358. Lefler, J.S., et al., Air pollution and mortality in a large, representative US cohort: multiple-pollutant analyses, and spatial and temporal decompositions. Environmental Health, 2019. 18(1): p. 1-11. 359. Tessum, C.W., et al., PM2. 5 polluters disproportionately and systemically affect people of color in the United States. Science Advances, 2021. 7(18): p. eabf4491. 360. Mohai, P. and R. Saha, Racial inequality in the distribution of hazardous waste: A national-level reassessment. Social problems, 2007. 54(3): p. 343-370. 361. Zwickl, K., M. Ash, and J.K. Boyce, Regional variation in environmental inequality: Industrial air toxics exposure in US cities. Ecological Economics, 2014. 107: p. 494-509. 362. Weaver, G.M. and W.J. Gauderman, Traffic-related pollutants: Exposure and Health effects among Hispanic children. American journal of epidemiology, 2018. 187(1): p. 45- 52. 363. Martin, J.A., et al., Births: Final Data for 2019. Natl Vital Stat Rep, 2021. 70(2): p. 1-51. 364. Alhusen, J.L., et al., Racial Discrimination and Adverse Birth Outcomes: An Integrative Review. J Midwifery Womens Health, 2016. 61(6): p. 707-720. 181 365. Howell, E.A., Reducing Disparities in Severe Maternal Morbidity and Mortality. Clin Obstet Gynecol, 2018. 61(2): p. 387-399. 366. Bekkar, B., et al., Association of air pollution and heat exposure with preterm birth, low birth weight, and stillbirth in the US: a systematic review. JAMA network open, 2020. 3(6): p. e208243-e208243. 367. Brauer, M., et al., A cohort study of traffic-related air pollution impacts on birth outcomes. Environ Health Perspect, 2008. 116(5): p. 680-6. 368. Fu, L., et al., The associations of air pollution exposure during pregnancy with fetal growth and anthropometric measurements at birth: a systematic review and meta- analysis. Environ Sci Pollut Res Int, 2019. 26(20): p. 20137-20147. 369. Ha, S., et al., The effects of air pollution on adverse birth outcomes. Environ Res, 2014. 134: p. 198-204. 370. Lee, P.C., et al., First trimester exposure to ambient air pollution, pregnancy complications and adverse birth outcomes in Allegheny County, PA. Matern Child Health J, 2013. 17(3): p. 545-55. 371. Rosa, M.J., et al., Prenatal exposure to PM2. 5 and birth weight: A pooled analysis from three North American longitudinal pregnancy cohort studies. Environment international, 2017. 107: p. 173-180. 372. Salam, M.T., et al., Birth outcomes and prenatal exposure to ozone, carbon monoxide, and particulate matter: results from the Children's Health Study. Environ Health Perspect, 2005. 113(11): p. 1638-44. 373. Stieb, D.M., et al., Ambient air pollution, birth weight and preterm birth: a systematic review and meta-analysis. Environmental research, 2012. 117: p. 100-111. 374. Blake, B.E. and S.E. Fenton, Early life exposure to per- and polyfluoroalkyl substances (PFAS) and latent health outcomes: A review including the placenta as a target tissue and possible driver of peri- and postnatal effects. Toxicology, 2020. 443: p. 152565. 375. Bové, H., et al., Ambient black carbon particles reach the fetal side of human placenta. Nature communications, 2019. 10(1): p. 1-7. 376. Perera, F. and J. Herbstman, Prenatal environmental exposures, epigenetics, and disease. Reproductive toxicology, 2011. 31(3): p. 363-373. 377. Saenen, N., et al., Air pollution-induced placental alterations: an interplay of oxidative stress, epigenetics, and the aging phenotype? Clinical epigenetics, 2019. 11(1): p. 1-14. 182 378. Barker, D.J., et al., Type 2 (non-insulin-dependent) diabetes mellitus, hypertension and hyperlipidaemia (syndrome X): relation to reduced fetal growth. Diabetologia, 1993. 36(1): p. 62-7. 379. Barker, D.J., et al., Trajectories of growth among children who have coronary events as adults. N Engl J Med, 2005. 353(17): p. 1802-9. 380. Gluckman, P.D., et al., Effect of in utero and early-life conditions on adult health and disease. N Engl J Med, 2008. 359(1): p. 61-73. 381. Aguilera, I., et al., Prenatal exposure to traffic-related air pollution and ultrasound measures of fetal growth in the INMA Sabadell cohort. Environmental health perspectives, 2010. 118(5): p. 705-711. 382. Iñiguez, C., et al., Prenatal exposure to traffic-related air pollution and fetal growth in a cohort of pregnant women. Occup Environ Med, 2012. 69(10): p. 736-44. 383. Iñiguez, C., et al., Prenatal Exposure to NO2 and Ultrasound Measures of Fetal Growth in the Spanish INMA Cohort. Environ Health Perspect, 2016. 124(2): p. 235-42. 384. Ritz, B., et al., Prenatal air pollution exposure and ultrasound measures of fetal growth in Los Angeles, California. Environ Res, 2014. 130: p. 7-13. 385. Wang, W., et al., Prenatal NO2 exposure and ultrasound measures of foetal growth: a prospective cohort study in Wuhan, China. Occupational and environmental medicine, 2017. 74(3): p. 204-210. 386. van den Hooven, E.H., et al., Air pollution exposure during pregnancy, ultrasound measures of fetal growth, and adverse birth outcomes: a prospective cohort study. Environ Health Perspect, 2012. 120(1): p. 150-6. 387. Clemens, T., S. Turner, and C. Dibben, Maternal exposure to ambient air pollution and fetal growth in North-East Scotland: A population-based study using routine ultrasound scans. Environ Int, 2017. 107: p. 216-226. 388. Hansen, C.A., A.G. Barnett, and G. Pritchard, The effect of ambient air pollution during early pregnancy on fetal ultrasonic measurements during mid-pregnancy. Environ Health Perspect, 2008. 116(3): p. 362-9. 389. Lamichhane, D.K., et al., Air pollution exposure during pregnancy and ultrasound and birth measures of fetal growth: A prospective cohort study in Korea. Sci Total Environ, 2018. 619-620: p. 834-841. 390. Zhao, N., et al., Effects of prenatal exposure to ambient air pollutant PM10 on ultrasound-measured fetal growth. Int J Epidemiol, 2018. 47(4): p. 1072-1081. 183 391. Cao, Z., et al., Maternal exposure to ambient fine particulate matter and fetal growth in Shanghai, China. Environmental health, 2019. 18(1): p. 1-8. 392. Leung, M., et al., Exposure to PM 2.5 during Pregnancy and Fetal Growth in Eastern Massachusetts, USA. Environmental Health Perspectives, 2022. 130(1): p. 017004. 393. Lin, L., et al., The associations of particulate matters with fetal growth in utero and birth weight: A birth cohort study in Beijing, China. Sci Total Environ, 2020. 709: p. 136246. 394. Najafi, M.L., et al., Preconception air pollution exposure and glucose tolerance in healthy pregnant women in a middle-income country. Environmental Health, 2020. 19(1): p. 1-10. 395. Rammah, A., K.W. Whitworth, and E. Symanski, Particle air pollution and gestational diabetes mellitus in Houston, Texas. Environmental Research, 2020. 190: p. 109988. 396. Robledo, C.A., et al., Preconception and early pregnancy air pollution exposures and risk of gestational diabetes mellitus. Environmental research, 2015. 137: p. 316-322. 397. Bastain, T.M., et al., Study Design, Protocol and Profile of the Maternal And Developmental Risks from Environmental and Social Stressors (MADRES) Pregnancy Cohort: a Prospective Cohort Study in Predominantly Low-Income Hispanic Women in Urban Los Angeles. BMC Pregnancy Childbirth, 2019. 19(1): p. 189. 398. Abatzoglou, J.T., Development of gridded surface meteorological data for ecological applications and modelling. International Journal of Climatology, 2013. 33(1): p. 121- 131. 399. Hadlock, F.P., et al., Estimation of fetal weight with the use of head, body, and femur measurements--a prospective study. Am J Obstet Gynecol, 1985. 151(3): p. 333-7. 400. Marsál, K., et al., Intrauterine growth curves based on ultrasonically estimated foetal weights. Acta Paediatr, 1996. 85(7): p. 843-8. 401. To, M. and S. Pereira, Routine Fetal Anomaly Scan. Twining's Textbook of Fetal Abnormalities E-Book, 2014. 402. Žaliūnas, B., et al., Fetal biometry: Relevance in obstetrical practice. Medicina, 2017. 53(6): p. 357-364. 403. Textor, J., et al., Robust causal inference using directed acyclic graphs: the R package ‘dagitty’. International journal of epidemiology, 2016. 45(6): p. 1887-1894. 404. Committee Opinion No 700: Methods for Estimating the Due Date. Obstet Gynecol, 2017. 129(5): p. e150-e154. 184 405. Gasparrini, A., Distributed lag linear and non-linear models in R: the package dlnm. Journal of statistical software, 2011. 43(8): p. 1. 406. Deguen, S., et al., Are the effects of air pollution on birth weight modified by infant sex and neighborhood socioeconomic deprivation? A multilevel analysis in Paris (France). Plos one, 2021. 16(4): p. e0247699. 407. Gale, C.R., et al., The influence of head growth in fetal life, infancy, and childhood on intelligence at the ages of 4 and 8 years. Pediatrics, 2006. 118(4): p. 1486-92. 408. Alkandari, F., et al., Fetal ultrasound measurements and associations with postnatal outcomes in infancy and childhood: a systematic review of an emerging literature. J Epidemiol Community Health, 2015. 69(1): p. 41-48. 409. Rückinger, S., et al., Growth in utero and body mass index at age 5 years in children of smoking and non-smoking mothers. Early Hum Dev, 2010. 86(12): p. 773-7. 410. Kannan, S., et al., Exposures to airborne particulate matter and adverse perinatal outcomes: a biologically plausible mechanistic framework for exploring potential effect modification by nutrition. Environmental health perspectives, 2006. 114(11): p. 1636- 1642. 411. Cushing, L., et al., Racial/Ethnic Disparities in Cumulative Environmental Health Impacts in California: Evidence From a Statewide Environmental Justice Screening Tool (CalEnviroScreen 1.1). Am J Public Health, 2015. 105(11): p. 2341-8. 412. Benmarhnia, T., et al., Decomposition analysis of Black–White disparities in birth outcomes: the relative contribution of air pollution and social factors in California. Environmental health perspectives, 2017. 125(10): p. 107003. 413. Rivera, N.Y.R., et al., Prenatal and early life exposure to particulate matter, environmental tobacco smoke and respiratory symptoms in Mexican children. Environmental Research, 2021. 192: p. 110365. 414. Shen, H.-N., et al., Maternal exposure to air pollutants and risk of gestational diabetes mellitus in Taiwan. International journal of environmental research and public health, 2017. 14(12): p. 1604. 415. Leung, M., et al., Bias due to Selection on Live Births in Studies of Environmental Exposures during Pregnancy: A Simulation Study. Environmental health perspectives, 2021. 129(4): p. 047001. 185 416. Barker, D. J.; Hales, C. N.; Fall, C. H.; Osmond, C.; Phipps, K.; Clark, P. M., Type 2 (non-insulin-dependent) diabetes mellitus, hypertension and hyperlipidaemia (syndrome X): relation to reduced fetal growth. Diabetologia 1993, 36, (1), 62-7. 417. Delpierre, C.; Lepeule, J.; Cordier, S.; Slama, R.; Heude, B.; Charles, M. A., [DOHaD: epidemiological researches]. Medecine sciences : M/S 2016, 32, (1), 21-6. 418. Martin, J. A.; Hamilton, B. E.; Osterman, M. J. K.; Driscoll, A. K., Births: Final Data for 2019. National vital statistics reports : from the Centers for Disease Control and Prevention, National Center for Health Statistics, National Vital Statistics System 2021, 70, (2), 1-51. 419. Mohai, P.; Pellow, D.; Roberts, J. T., Environmental Justice. Annual Review of Environment and Resources 2009, 34, (1), 405-430. 420. Liévanos, R. S., Retooling CalEnviroScreen: Cumulative pollution burden and race-based environmental health vulnerabilities in California. International Journal of Environmental Research and Public Health 2018, 15, (4), 762. 421. Solomon, G. M.; Morello-Frosch, R.; Zeise, L.; Faust, J. B., Cumulative environmental impacts: science and policy to protect communities. Annual Review of Public Health 2016, 37, 83-96. 422. Malits, J.; Blustein, J.; Trasande, L.; Attina, T. M., Perfluorooctanoic acid and low birth weight: estimates of US attributable burden and economic costs from 2003 through 2014. International Journal of Hygiene and Environmental Health 2018, 221, (2), 269-275. 423. Cave, B. PFAS Update: State-by-State Regulation of PFAS Substances in Drinking Water. https://www.jdsupra.com/legalnews/pfas-update-state-by-state-regulation- 4639985/ 424. OEHHA, C., The Proposition 65 List. In OEHHA. https://oehha. ca. gov/proposition- 65/proposition-65-list: 2021. 425. Information, C. L. Assembly Bill No. 1200. https://leginfo.legislature.ca.gov/faces/billNavClient.xhtml?bill_id=202120220AB1200 426. Planning, U. S. E. P. A. O. o. A. Q.; Standards, Review of the national ambient air quality standards for particulate matter: Policy assessment of scientific and technical information. DIANE Publishing: 1996. 427. Table, N., EPA. In 2016. 428. Trasande, L.; Malecha, P.; Attina, T. M., Particulate matter exposure and preterm birth: estimates of US attributable burden and economic costs. Environmental health perspectives 2016, 124, (12), 1913-1918. 186 429. Gale, C. R.; O'Callaghan, F. J.; Bredow, M.; Martyn, C. N.; Parents, A. L. S. o.; Team, C. S., The influence of head growth in fetal life, infancy, and childhood on intelligence at the ages of 4 and 8 years. Pediatrics 2006, 118, (4), 1486-1492. 430. Mogensen, U. B.; Grandjean, P.; Nielsen, F.; Weihe, P.; Budtz-Jørgensen, E., Breastfeeding as an Exposure Pathway for Perfluorinated Alkylates. Environmental science & technology 2015, 49, (17), 10466-73. 431. Liu, Y.; Li, N.; Papandonatos, G. D.; Calafat, A. M.; Eaton, C. B.; Kelsey, K. T.; Chen, A.; Lanphear, B. P.; Cecil, K. M.; Kalkwarf, H. J.; Yolton, K.; Braun, J. M., Exposure to Per- and Polyfluoroalkyl Substances and Adiposity at Age 12 Years: Evaluating Periods of Susceptibility. Environmental science & technology 2020, 54, (24), 16039-16049. 432. Qi, W.; Clark, J. M.; Timme-Laragy, A. R.; Park, Y., Per- and Polyfluoroalkyl Substances and Obesity, Type 2 Diabetes and Non-alcoholic Fatty Liver Disease: A Review of Epidemiologic Findings. Toxicological and environmental chemistry 2020, 102, (1-4), 1-36. 433. Bello, G. A.; Arora, M.; Austin, C.; Horton, M. K.; Wright, R. O.; Gennings, C., Extending the Distributed Lag Model framework to handle chemical mixtures. Environmental research 2017, 156, 253-264. 434. Liu, S. H.; Bobb, J. F.; Claus Henn, B.; Schnaas, L.; Tellez‐Rojo, M. M.; Gennings, C.; Arora, M.; Wright, R. O.; Coull, B. A.; Wand, M. P., Modeling the health effects of time‐ varying complex environmental mixtures: Mean field variational Bayes for lagged kernel machine regression. Environmetrics 2018, 29, (4), e2504. 435. Liu, S. H.; Bobb, J. F.; Lee, K. H.; Gennings, C.; Claus Henn, B.; Bellinger, D.; Austin, C.; Schnaas, L.; Tellez-Rojo, M. M.; Hu, H., Lagged kernel machine regression for identifying time windows of susceptibility to exposures of complex mixtures. Biostatistics 2018, 19, (3), 325-341. 436. Tanner, E.; Lee, A.; Colicino, E., Environmental mixtures and children’s health: identifying appropriate statistical approaches. Current opinion in pediatrics 2020, 32, (2), 315. 437. Wilson, A.; Hsu, H. H. L.; Chiu, Y. H. M.; Wright, R. O.; Wright, R. J.; Coull, B. A. In Bayesian Kernel Machine Regression Distributed Lag Models, ISEE Conference Abstracts, 2021; 2021. 438. Morales-McDevitt, M. E.; Becanova, J.; Blum, A.; Bruton, T. A.; Vojta, S.; Woodward, M.; Lohmann, R., The air that we breathe: Neutral and volatile PFAS in indoor air. Environmental Science & Technology Letters 2021, 8, (10), 897-902.

Abstract (if available)

Abstract The gestational period of the life course is a particularly susceptible period to the effects of environmental exposures. Such exposures can negatively impact fetal development during pregnancy and lead to increased disease morbidity in later childhood and adulthood. Studies have documented that exposure to chemical contaminants differs by racial and ethnic groups, as does the likelihood of having an infant be born with low birthweight (<2,500 grams). Additionally, populations facing health disparities tend to live in communities that house multiple sources of pollution due to a variety of factors such as the proximity to major traffic sources and industrial facilities. Beyond higher exposure levels, these communities may be more vulnerable to health effects from these pollutants due to the compounded challenge of structural inequities related to healthcare access, economic disadvantage, and social stressors. This dissertation investigates the influence of two environmental exposures possibly experienced during pregnancy and their impacts on measures of fetal growth within the Maternal And Developmental Risks from Environmental and Social Stressors (MADRES) pregnancy cohort—a cohort of predominately low-income Hispanic participants in urban Los Angeles, California. We investigate per-and polyfluoroalkyl substances (PFAS) measured in prenatal blood serum and the association with infant birthweight and in utero fetal growth trajectories. We demonstrate that prenatal exposure to perfluorooctanoic acid (PFOA) is associated with fetal growth in two ways: 1) by reducing infant birthweight and 2) by impacting in utero fetal growth with reductions in head circumference and biparietal diameter over pregnancy. Notably, we provide novel evidence that the association of PFOA exposure on in utero fetal growth may be modified by maternal perceived stress experienced during pregnancy. This dissertation also examines how ambient air pollution experienced in the three months prior to conception, as well as across gestation, influences fetal growth measured in the third trimester of pregnancy within MADRES participants. This is one of the first studies to assess these associations using distributed lag modeling (DLM) techniques, which allowed critical windows of exposure to be identified with fine temporal resolution. We found that particulate matter with an aerodynamic diameter of 2.5 μm or less (PM2.5) during weeks four to sixteen of gestation was associated with significantly lower estimated fetal weight measured in the third trimester of pregnancy and PM2.5 exposure during weeks one to twenty-three weeks of pregnancy was associated with significantly decreased fetal abdominal circumference. Moreover, particulate matter with an aerodynamic diameter of 10 μm or less (PM10) exposure during weeks six to fifteen of pregnancy was associated with significantly decreased third trimester fetal abdominal circumference. Overall, we provide strong evidence that two different classes of environmental chemicals produce negative effects on fetal growth within a health disparities population. These findings have the potential to influence future policy decisions about harmful contaminants to help mitigate the future effects of exposure on fetal growth. Further research should explore mixture methods approaches for these exposures experienced concurrently and elucidate which specific components of particulate matter influence adverse fetal growth across pregnancy.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Prenatal exposure to organophosphate esters and early neurodevelopment

PDF

Personal exposure to particulate matter PM2.5 sources during pregnancy and birthweight

PDF

Associations of cumulative pollution burden and environmental health vulnerabilities with gestational weight gain in a cohort of predominantly low-income Hispanic women

PDF

Personal PM2.5 exposure during pregnancy in an environmental health disparities population

PDF

Sensor-based mobile health approaches for personal air pollution and pediatric asthma studies

PDF

Associations between ambient air pollution and hypertensive disorders of pregnancy

PDF

Prenatal air pollution exposure, newborn DNA methylation, and childhood respiratory health

PDF

Airway inflammation and respiratory health in the Southern California children's health study

PDF

Air pollution, mitochondrial function, and growth in children

PDF

Spatial analysis of PM₂.₅ air pollution in association with hospital admissions in California

PDF

Early life air pollution exposure, gestational diabetes mellitus, and autism spectrum disorder

PDF

A cohort study of air-pollution and childhood obesity incidence

PDF

Air pollution and breast cancer survival in California teachers: using address histories and individual-level data

PDF

Spatial modeling of non-tailpipe emissions and its association with children's lung function

PDF

The interplay between tobacco exposure and polygenic risk score for growth on birthweight and childhood acute lymphoblastic leukemia

PDF

Assessing the impact of air pollution on adverse birth outcomes in a low resource setting

PDF

Associations of ambient air pollution exposures with perceived stress in the MADRES cohort

PDF

Association of maternal and environmental factors with infant feeding behaviors in a birth cohort study

PDF

Chronic eye disease epidemiology in the multiethnic ophthalmology cohorts of California study

PDF

Native American ancestry among Hispanic Whites is associated with higher risk of childhood obesity: a longitudinal analysis of Children’s Health Study data

Asset Metadata

Creator Peterson, Alicia Kristin (author)

Core Title Prenatal environmental exposures and fetal growth in the MADRES cohort

School Keck School of Medicine

Degree Doctor of Philosophy

Degree Program Epidemiology

Degree Conferral Date 2022-08

Publication Date 06/29/2022

Defense Date 05/04/2022

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag ambient air pollution,birthweight,fetal growth,health disparities,OAI-PMH Harvest,particulate matter,PFAS,PFOA,Pregnancy

Format application/pdf (imt)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Breton, Carrie (committee chair), Bastain, Theresa (committee member), Eckel, Sandrah (committee member), Grubbs, Brendan (committee member), Habre, Rima (committee member)

Creator Email aliciakpeterson93@gmail.com,aliciap@usc.edu

Permanent Link (DOI) https://doi.org/10.25549/usctheses-oUC111352078

Unique identifier UC111352078

Legacy Identifier etd-PetersonAl-10798

Document Type Dissertation

Format application/pdf (imt)

Rights Peterson, Alicia Kristin

Type texts

Source 20220706-usctheses-batch-950 (batch), University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the author, as the original true and official version of the work, but does not grant the reader permission to use the work if the desired use is covered by copyright. It is the author, as rights holder, who must provide use permission if such use is covered by copyright. The original signature page accompanying the original submission of the work to the USC Libraries is retained by the USC Libraries and a copy of it may be obtained by authorized requesters contacting the repository e-mail address given.

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA

Repository Email cisadmin@lib.usc.edu