PHYSIOLOGICAL, ENVIRONMENTAL, AND GENETIC INFLUENCES ON IRON HOMEOSTASIS A Dissertation Presented to the Faculty of the Graduate School of Cornell University in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy by Wanhui Kang August 2021 © 2021 Wanhui Kang PHYSIOLOGICAL, ENVIRONMENTAL, AND GENETIC INFLUENCES ON IRON HOMEOSTASIS Wanhui Kang, Ph.D. Cornell University 2021 Iron deficiency is the most widespread micronutrient deficiency globally, while elevated iron stores are also associated with an increased risk of adverse health outcomes. Iron homeostasis must be strictly maintained to ensure that adequate iron is available for metabolic functions and to avoid the toxicity that results from iron overload. The overall goal of this research was to investigate selected physiological (pregnancy), environmental (parasitic infections and iron supplementation) and genetic (ethnic differences in iron status) influences on iron homeostasis. Pregnancy places a significant strain on iron homeostasis. Surprisingly, there are very few longitudinal studies of hemoglobin concentrations across gestation and the norms that are used to define gestational anemia are based on limited evidence. To address a physiological determinant of iron homeostasis, this research provided a large normative analysis of determinants of maternal hemoglobin concentrations and identified factors associated with risk of maternal anemia in healthy U.S. pregnant women. Anemia was prevalent in otherwise healthy U.S. pregnant women even with the current universal iron supplementation recommendations. The existing CDC anemia cutoffs may overestimate the prevalence of anemia in late pregnancy based on our findings that were obtained in a significantly larger and more racially diverse population. Risk factors for anemia were identified as Black race, being underweight at entry into pregnancy, being multiparous or carrying a multiple birth pregnancy. Multiple environmental factors can influence iron status and risk of iron deficiency and iron deficiency anemia. Parasitic infections are one example of a common environmental factor that can cause anemia. To obtain more detailed information on the impact of parasitic infections and iron supplementation, stable iron isotopes were used to investigate iron absorption and change in erythrocyte iron isotope enrichment in a group of toddlers residing in a malaria and helminth endemic environment. Toddlers who received iron supplementation exhibited greater loss of erythrocyte iron isotope enrichment over time, which may be indicative of an increase in the size of the erythrocyte pool and increased erythropoietic drive due to the provision of iron among anemic or helminth infected toddlers. This study also summarized published literature on genetic determinants of iron status, characterized iron status in different ethnic populations, and statistically evaluated differences in iron status as a function of ethnicity using published data. A review of the literature identified multiple iron-related and non-iron related genes that have been associated with iron status. For the first time, these published data were used to statistically explore differences in iron status among various ethnic populations. East Asians had higher iron status compared to other racial groups yet this group also has the lowest prevalence of HFE mutations that are commonly associated with elevated iron stores. Future research using multidisciplinary approaches that consider physiological, environmental, and genetic factors is needed to better predict iron status at the individual and population level and to help develop the most effective interventions to maintain adequate iron status. BIOGRAPHICAL SKETCH Wanhui Kang was born in Taiyuan, China on November 9th,1993 to Fang Wang and Hongbing Kang. Wanhui graduated from Taiyuan No. Five Middle School in 2012 and was accepted to China Agriculture University in Beijing, where she majored in Food Quality and Safety. In her junior year, she was selected by China Agricultural University to study abroad at Cornell University, and she changed her major to Nutritional Sciences because of her passion for human nutrition. In June 2017, Wanhui graduated from Cornell University summa cum laude as Banner Bearer and from China Agricultural University and obtained dual bachelor’s degrees. After graduating, Wanhui entered the doctoral program in nutrition in the Division of Nutritional Sciences at Cornell University, where she joined Dr. Kimberly O’Brien’s research group. Her dissertation focuses on understanding the impact of pregnancy, environmental exposures, and ancestry on iron homeostasis. iii DEDICATION This dissertation is dedicated to my mother, 王芳 (Fang Wang) and my grandparents 王文华 (Wenhua Wang) and 连国英 (Guoying Lian). I love you forever. iv ACKNOWLEDGEMENTS I feel extremely grateful to many people who have accompanied me on this academic journey and who have helped me accomplish my goals. First and foremost, I would like to express my sincere gratitude to my advisor, Dr. Kimberly O’Brien, for her warmth, support and enormous amount of time dedicated towards my academic and professional training. I have learned so many invaluable things from Kimberly in many aspects, not only attitudes and passion for science and research, qualities of being an excellent researcher and mentor, but also the attitudes towards life and challenges. I am grateful for her kindness and support during difficult times and for her help in keeping me engaged with our laboratory team, so I did not feel isolated due to the travel restrictions caused by the COVID-19 pandemic that prevented me from returning to the United States. I always feel fortunate and thankful that I have had the opportunity to enter the academic field under Kimberly’s guidance. I would also like to thank my committee members, Dr. Andrew Clark, Dr. Bethany Cummings, Dr. Elizabeth Johnson and Dr. Zhenglong Gu, who supported me and provided generous advice on academic and professional development. I am also grateful to Francoise Vermeylen, Lynn Johnson and Erika Mudrak from the Cornell Statistical Consulting Unit who have helped me a lot with their expertise on statistics and always been very patient. I would like to thank Dr. Eva Pressman, Carrie Irvine and Andrea Hart from the University of Rochester / Strong Memorial Hospital who helped with the obstetric data extraction and made the third dissertation aim possible. I am appreciative for the mothers and toddlers who participated in the Pemba study, for v all the participants who participated in the FeGenes study, and for the funding sources that made these projects possible. I am also grateful for the graduate students that I have worked with in the O’Brien lab, Dr. Cora Best, Dr. Katherine Delaney, Luisa Castillo and Alexa Barad. They provided me with intellectual feedback and offered me support when I became stranded in China. I warmly appreciate the support from my friends in DNS, especially that provided by Drs’ Laura Pompano, Yuan Ru, Jiayi Xu, Yingyi Zhong, Yue Qu, and Yiqin Wang, for their help and the time they spent providing me with constructive feedback about research and life. Finally, I would like to thank my family for their love and support. They always have my back and support every step of my life. I could not have achieved anything without their love, support, and protection. vi TABLE OF CONTENTS Biographical sketch …………………………………………………………………. iii Dedication …………………………………………………………………………….iv Acknowledgements …………………………………………………………...……….v Table of contents ……………………………………………………………………..vii List of Figures ………………………………………………………………………...ix List of Tables ………………………………………………………………...………..x List of Abbreviations …………………………………………………………...…….xi Chapter 1: Introduction ……………………………………………………………... 1 Specific Aims ………………………………………………………………… 2 Background and Significance………………………………………………….6 I. Iron homeostasis…………………………………………………………...6 II. Physiological influences on iron homeostasis……………………..……..16 III. Impact of environmental factors on iron homeostasis……………………19 IV. Genetic influences on iron homeostasis ………………………………….23 V. Significance and Summary…………………………..…………………...25 References ……………………………………………………………............28 Chapter 2: Prevalence and determinants of anemia in healthy US pregnant women..45 Abstract ……………………………………………………………………....46 Introduction …………………………………………………………………..48 Methods ……………………………………………………………………....50 Results ………………………………………………………………………..53 Discussion …………………………………………………………………....71 References ………………………………………………………...………….77 Chapter 3: Iron supplementation in anemic Zanzibari toddlers is associated with greater loss in erythrocyte iron isotope enrichment ………………………………….80 Abstract ……………………………………………………………………....81 Introduction …………………………………………………………………. 83 Subjects and Methods ………………………………………………………..84 Results ………………………………………………………………………. 89 Discussion …………………………………………………………………..100 References …………………………………………………………………. 105 Chapter 4: Ethnic differences in iron status………………………….. ………...…110 Abstract ……………………………………………………………………. 111 Introduction ……………………………………………………...………… 112 Current Status of Knowledge………………………………………….…….113 Iron physiology…………………………………………………………...…113 Genetic variants of iron metabolism…………………………...……………117 Ethnic differences in iron status…………………………………...……...…130 Statistical evaluation of population-based data on iron status ……..……….135 vii Conclusions…………………………...……………………………...……...139 References …………………………………………………………………..141 Chapter 5: Summary and conclusions …………………………………………..…168 Summary ………………………………………………………………..…..169 Future Directions …………………………………………………………... 171 Appendix……………………………………………………………………………172 References……………………………………………………..…………… 179 viii LIST OF FIGURES Chapter 1 Figure 1.1 Three major factors that influence iron homeostasis ……………….....3 Figure 1.2 Three major factors that influence iron homeostasis and corresponding dissertation chapters devoted to each of these key factors.….………. 25 Chapter 2 Figure 2.1 Flow chart of the study population identified from the retrospective medical chart review………………………..………………………...54 Figure 2.2 Hemoglobin distributions across gestation……………………..…….59 Figure 2.3 Hemoglobin concentrations across gestation…………..……….....….60 Figure 2.4 Self-reported racial groups and anemia prevalence ………………….64 Figure 2.5 Global distributions of hemoglobin concentration for pregnant and non- pregnant women in 2011 presented by Steven et al ………..……..….72 Chapter 3 Figure 3.1 Flow chart of toddlers enrolled in the study …………………………89 Figure 3.2 Mean change in RBC iron DAE (D84-D14) in toddlers as a function of sex, iron status or parasitic burden……………………………............98 Chapter 4 Figure 4.1 The hepatic hepcidin signaling pathway and common genetic mutations in this signaling pathway…………………………………………….115 ix LIST OF TABLES Chapter 1 Table 1.1 Biochemical iron status indicators ………………….…….. ………...14 Chapter 2: Table 2.1 Characteristics of pregnant women ………………….…….. ………..56 Table 2.2 Hemoglobin concentrations and anemia/elevated hemoglobin prevalence across gestation ………………….…….. ………………..60 Table 2.3 Mean hemoglobin concentrations as a function of key study variables………………………………………………………………62 Table 2.4 Prevalence of anemia as a function of key study variables…………...65 Table 2.5 Prevalence of elevated hemoglobin concentrations (> 13.0 g/dL) as a function of key study variables………………………………….........66 Table 2.6 Determinants of anemia………………………………………………69 Chapter 3 Table 3.1 General characteristics of the 71 toddlers enrolled in the study…...…90 Table 3.2 Iron status in anemic toddlers that received, or did not receive, iron supplementation on D14……………………………………………...91 Table 3.3 Iron status as a function of malaria parasitemia…..……………. ……94 Table 3.4 Factors associated with red blood cell iron incorporation (%) at D14..96 Chapter 4 Table 4.1 Mean reported iron status indicators in different ethnic groups ……137 Table 4.2 Comparison of iron status indicators between comparison ethnic groups and reference ethnic groups ………………………………………...138 Appendix Appendix Table 1 Commonly utilized biochemical indicators of iron status…….173 Appendix Table 2 Genetic variants shown to be associated with biomarkers of iron status in healthy populations stratified by ethnic group……….175 Appendix Table 3 Summary of the literature on iron status in different ethnic population………………………………………………..…….178 x LIST OF ABBREVIATIONS CRP C-Reactive Protein DMT1 Divalent Metal Transporter 1 EPO Erythropoietin ERFE Erythroferrone FPN Ferroportin Hb Hemoglobin HIF Hypoxia-Inducible Factor HJV Hemojuvelin ID Iron Deficiency IDA Iron Deficiency Anemia IL-6 Interleukin 6 IRE Iron Responsive Element IRP Iron Regulatory Protein LBW Low Birth Weight MT-2 Matriptase-2 NTBI Non-Transferrin Bound Iron RBC Red Blood Cell SF Serum Ferritin SGA Small for Gestational Age sTfR Soluble Transferrin Receptor TSAT Transferrin Saturation TBI Total Body Iron TfR Transferrin Receptor TIMS Thermal Ionization Mass Spectrometry TMPRSS6 Transmembrane Serine Protease 6 ZIP 14 Zrt/Irt-like protein-4 xi CHAPTER 1 INTRODUCTION 1 Specific Aims Iron is an essential trace element in almost all living organisms as it is required for a variety of metabolic and catabolic processes (1). Iron deficiency (ID) is the most widespread micronutrient deficiency, affecting an estimated 2.5 billion people worldwide (2,3). At the other extreme, iron excess is also detrimental to health as this leads to the generation of free radicals that can cause oxidative damage. Human epidemiological studies have documented that elevated iron stores are associated with an increased risk of developing type II diabetes (independent of inflammation) (4,5), cardiovascular disease (6,7), liver fibrosis, cancer (8–10), and neurodegenerative diseases (11,12). Iron balance must be strictly maintained to ensure that adequate iron is available for metabolic functions and to avoid the toxicity that results from iron overload. Unlike other essential micronutrients, iron is unique because there is no regulatable excretory pathway to eliminate excess iron once it has been absorbed. Thus, iron homeostasis in humans depends almost exclusively on the tight regulation of dietary iron absorption by enterocytes. There are three major factors that influence the regulation of iron homeostasis, namely environment, genetics and physiological state (Figure 1.1). 2 Figure 1.1 Three major factors that influence iron homeostasis. Physiological stages such as infancy, childhood, adolescence and pregnancy have a significant impact on iron homeostasis. Iron requirements change because of physical growth during childhood, increased physiological losses of iron in women due to menstrual blood loss, and due to the increased maternal and fetal iron demands of pregnancy. During pregnancy, physiologic iron demands increase substantially to support fetoplacental development and maternal adaptations to pregnancy.(13) Extremes of iron status during pregnancy (both maternal anemia and high hemoglobin (Hb) concentrations) have been reported to increase the risks of unfavorable birth outcomes such as small-for-gestational-age (SGA), preterm birth, and perinatal death (14). However limited normative data in healthy U.S. women with uncomplicated pregnancies are available to establish cut-offs for maternal Hb concentrations across pregnancy. Many environmental factors influence iron status. Among these, parasitic infections caused by malaria or geohelminths can lead to inflammation-mediated increases in serum hepcidin. This chronic increase in hepcidin leads to a sequestration of iron in macrophages and decreased intestinal dietary iron absorption, thereby increasing the risk of anemia (15). Another 3 important environmental factor that impacts iron status is iron intake from diet and/or supplements. Adequate dietary or supplemental iron can help ameliorate ID and iron deficiency anemia (IDA). However, there are situations where administration of supplemental iron can be detrimental. One example is the use of supplemental iron for children residing in malaria- endemic areas (16). Studies conducted in malaria-endemic areas have found that untargeted oral iron supplementation increases the risk of contracting infectious diseases and led to increased mortality among preschool children (17,18). Evidence addressing mechanisms whereby helminth and malarial infections interact to impact iron absorption and response to supplemental iron treatment in affected children are not fully understood. To date little is known about iron absorption and utilization in children with mild chronic parasitic infections (19). In addition, the effects of oral iron supplementation on iron utilization and long-term iron metabolism in children residing in parasite-endemic areas is unknown. A third factor that influences the regulation of iron homeostasis is genetic background. To date only about 30% of the interindividual variability in iron absorption can be captured using measures of systemic iron status or by iron regulatory hormones such as hepcidin (20–24). Variability in iron status and absorption in healthy populations may be attributed to genetic factors. While there are well known ethnic differences in risk of iron overload, no studies to date have summarized published data on genetic variants found to be associated with iron metabolism across different ethnic groups nor have published data evaluated iron status as a function of ethnicity. 4 To address existing gaps in knowledge on iron homeostasis, the goal of this dissertation is to improve our understanding of physiological, environmental and genetic influences on iron homeostasis. This dissertation will address the following specific aims and hypotheses: Specific Aims and Hypothesis Specific Aim 1: To characterize maternal Hb distributions and the prevalence of anemia in a large cohort of pregnant women and to identify factors associated with risk of anemia in this cohort. Hypothesis: The 5th percentile of the Hb distribution used currently by the CDC will be closer to the 50th percentile of the Hb distribution in a modern pregnancy cohort. Mean Hb concentrations per trimester will be impacted by race/ancestry. The prevalence of anemia will be highest in the third trimester of pregnancy and highest in Black women, while the prevalence of high Hb concentrations will be highest in the first trimester and among Asian and White women. Specific Aim 2: To utilize a stable iron isotope approach to identify the impact of parasitic infections and iron supplementation on iron incorporation into red blood cells (RBC) and to identify variables associated with the change in RBC iron isotope enrichment over a 3-month period among a group of toddlers residing in a malaria and helminth endemic environment. Hypothesis: Iron status and iron supplementation will be negatively associated with iron incorporation into RBC’s. Toddlers with parasitic infections will exhibit higher RBC Fe incorporation and a greater loss of RBC iron isotope enrichment post-dosing. 5 Specific Aim 3: To review genetic determinants of iron status and to characterize iron status in different ethnic populations and to undertake a statistical evaluation of possible differences in iron status as a function of ethnicity using published data. Hypothesis: Both iron- and non-iron related genes or single-nucleotide polymorphisms (SNPs) will be associated with variations in iron traits among healthy populations. The frequencies of certain mutations or SNPs will differ as a function of ethnicity. East Asian and White populations will have the highest risk of elevated iron status while Black populations will have the highest risk of low iron status. Background and Significance I. Iron homeostasis a. General Iron Physiology Iron is an essential trace element involved in numerous metabolic processes, including oxygen transport and utilization, cellular proliferation, DNA synthesis, neurotransmitter synthesis, and energy production (1,25). Major organs or systems involved in iron metabolism include the small intestine, bone marrow, liver, and reticuloendothelial system. Four key cell types play key roles in regulation of iron homeostasis including the enterocytes, erythrocytic precursors, macrophages, and hepatocytes. Iron is absorbed into the blood stream across the enterocytes, specifically those of the proximal small intestine. Absorbed iron then binds to its specific systemic iron transport protein, transferrin. The majority of transferrin bound iron is delivered to erythrocyte precursors in the bone marrow to be incorporated into Hb in newly 6 synthesized RBCs (26). The distribution of total body iron is as follows: approximately two thirds of total body iron is contained within the circulating erythrocytes as heme. Heme consists of an iron ion complexed within a porphyrin ring. This complex is an indispensable component of Hb. Approximately 10-25% of body iron is stored as ferritin primarily in the liver and reticuloendothelial macrophages. This iron reserve can be mobilized when dietary iron absorption cannot meet iron demands. The remaining 5-15% of body iron is present within the heme component of myoglobin in muscle tissues and in a variety of enzymes involved in oxidative metabolism (26–29). Unlike other essential micronutrients, iron is unique because there are no regulatable excretory pathways to eliminate excess iron once it has been absorbed. Thus, iron homeostasis in humans depends almost exclusively on the tight regulation of dietary iron absorption by enterocytes. There are two forms of iron ingested from dietary sources: heme iron and non-heme iron, and these two forms of iron are absorbed using different mechanisms. Absorption of non- heme iron (from both animal and plant sources) is regulated in response to iron status, whereas absorption of heme iron (from meat based dietary sources) is not tightly regulated in relation to iron status. Non-heme iron accounts for most of the dietary iron ingested primarily as ferric iron (Fe3+) or as ferrous iron (Fe2+) or in plant sources, animal tissues, or from iron fortificants that have been added to food during processing. Dietary ferric iron must be first reduced to ferrous iron (Fe2+) by a ferrireductase, duodenal cytochrome B, before it can be transported into the enterocyte by the divalent metal transporter 1 (DMT1) protein. Once ferrous iron (Fe2+) enters the enterocyte, it can either be stored as ferritin or exported across the basolateral membrane of 7 the enterocyte via iron export protein, ferroportin (FPN). As iron is exported across the basolateral membrane of the enterocyte, the ferrioxidase, hephaestin, oxidizes ferrous iron (Fe2+) to ferric iron (Fe3+) in order for iron to bind to serum transferrin to be transported throughout the blood stream to other body tissues. Transferrin-bound ferric iron is taken up into target cells that express transferrin receptor 1 (TfR1) by receptor mediated endocytosis. Cellular expression of TfR1 is highest in erythroid tissue to support the iron demands of erythropoiesis. Under normal circumstances, approximately 30% of the iron-binding sites in the plasma transferrin pool are occupied (30,31). This value is expressed as a percentage called transferrin saturation (TSAT), i.e. the ratio of serum iron/total iron binding capacity. When TSAT exceeds 45%, iron begins to circulate free or bound to low molecular weight molecules (citrate, albumin) generating a potentially toxic iron species known as non-transferrin bound iron (NTBI). The pancreatic cells and hepatocytes have a high capacity to internalize NTBI via ZIP14, while cardiomyocytes take up NTBI using L-type calcium channel and T-type calcium channel (32–34). Excess NTBI can increase the intracellular labile iron pool causing the generation of reactive oxygen species and free radicals that can cause oxidative damage particularly in organs that express ZIP 14 (liver and pancreas) or in the heart. Only 1-2 mg of absorbed iron per day is needed to offset the typical amounts of daily iron lost from cell desquamation. The majority of iron needed to meet the high (20-25 mg/day) iron demands of erythropoiesis is obtained from the daily catabolism of senescent RBC’s (30,35). The heme iron that is released in the macrophages of the liver, spleen, and bone marrow is crucial for maintaining physiological erythropoiesis. In general, disruption of all pathways 8 mentioned above, especially iron absorption, would result in abnormal iron metabolism and utilization, ultimately leading to adverse or even detrimental health consequences. b. Regulation of Iron Homeostasis Humans have evolved sophisticated mechanisms to regulate non-heme iron absorption. Iron absorption can be regulated at the cellular level by specific iron trafficking proteins and is regulated systematically by several hormones. The enterocyte can sense cellular iron needs through the “iron regulatory proteins (IRP) and iron response elements (IRE)” system and respond by altering expressions of iron transporters involved in cellular iron uptake and egress. In response to low iron concentrations in the labile iron pool, IRP can either bind to specific IRE located in the 3’ or 5’ untranslated regions of mRNAs that encode for key iron transport proteins, allowing them to block or stabilize their transcription. For example, when little iron is available in the labile iron pool level, the IRP1 and IRP2 in the enterocytes would bind to 5’ region of the mRNA to block the translation of iron export and sequestering proteins such as FPN and ferritin; IRP’s can also bind to 3’ end of mRNA to stabilize the translation of genes that code for iron uptake proteins, such as DMT1 and TfR. In iron-replete cells, Fe/S clusters convert IRP1 into cytosolic aconitase, while IRP2 undergoes iron-dependent proteasomal degradation (36). Iron absorption and iron homeostasis can also be regulated systemically via three iron regulatory hormones: erythropoietin (EPO), hepcidin, and erythroferrone (ERFE). The first of these three hormones to be identified was erythropoietin (EPO), which was found in the 1950’s (37). Erythropoietin (EPO) is produced mainly by the kidneys in response to cellular hypoxia and when EPO binds to the EPO receptor in bone marrow this stimulates RBC production. 9 Decreased oxygen concentrations are detected by the kidneys which serves to stimulate EPO production. This process is mediated by hypoxia-inducible factors (HIF’s), which are transcription factors that respond to reduced tissue oxygen levels (38). Hepcidin is a small cysteine rich peptide hormone discovered in 2000 (1,39). Hepcidin is the key regulator of iron absorption and homeostasis. It is synthesized and secreted by the liver in response to inflammation, iron stores, oxygen tension, or erythropoietic demands. This hormone binds to FPN, causing it to be internalized, phosphorylated, ubiquitinated, and degraded within lysosomes. When present at higher concentrations, hepcidin can directly block FPN without endocytosis (40). This interaction limits iron egress from cells into the circulation by reducing absorption of dietary iron, recycling of iron from senescent erythrocytes, and release of iron from body iron reserves. Iron regulated hepcidin expression is mediated primarily through the BMP-SMAD pathway. Accumulated iron stores in the liver upregulate BMP ligands, which along with the constitutively active BMP type II receptors phosphorylate and activate BMP type I receptors. This activation results in the phosphorylation of SMAD1/5/8, a mechanism possibly mediated by endofin (41). Phosphorylated SMAD1/5/8 associates with cytoplasmic SMAD4 (42) and this complex translocates to the nucleus where it binds to BMP responsive elements in the hepcidin promoter, resulting in the activation of hepcidin transcription (43). Although several BMP ligands are expressed in the liver, only two of them (BMP2 and BMP6) have been shown to play a role in hepcidin expression (42). Membrane-bound hemojuvelin (HJV) functions as an essential BMP co-receptor that binds selectively to BMP6 ligands, which may be facilitated by 10 neogenin (44), to enhance SMAD phosphorylation. Whereas the inactive form of HJV, soluble HJV, may down-regulate hepcidin in a competitive way interfering with BMP signaling (45). Another key player in the regulation of the BMP-SMAD signaling pathway is the major histocompatibility complex class I like protein HFE. Under high concentrations of saturated transferrin, diferric-transferrin binds to TFR1 and displaces HFE from TFR1. HFE is then able to form a complex with TFR2 and possibly HJV to promote BMP-SMAD signaling (44). HFE also interacts with and stabilizes ALK3 to prevent its degradation, subsequently promoting BMP- SMAD signaling and hepcidin transcription (46). Conversely, under conditions of iron deficiency, the transmembrane serine protease 6 (TMPRSS6 or matriptase-2, MT2) exhibits higher stability (47) and cleaves HJV into its inactive form, which results in the inactivation of the BMP-SMAD signaling pathway and downregulation of hepcidin transcription. In addition, the immunophilin FKBP12 inhibits the BMP-SMAD pathway through its interaction with ALK2 to dampen hepcidin expression (48). Increased erythropoietic demands result in the downregulation of hepcidin expression via the function of EFRE (49,50). ERFE is a physiologic erythroid regulator mainly expressed in EPO-stimulated erythroblasts in the bone marrow (51). Decreased oxygen concentrations are detected by the kidneys via the HIF system, which stimulates the production and secretion of EPO from the kidneys. Elevated EPO then upregulates ERFE likely via the JAK2-STAT5 pathway (51) and ERFE suppresses hepcidin expression by sequestering BMP2/6 and inhibiting the hepatic BMP-SMAD signaling pathway (51–55). The specific molecular mechanism of action of ERFE on hepcidin is evolving and further studies are needed to solidify and expand 11 what is currently known. Moreover, under hypoxic conditions, decreased oxygen tension leads to decreased hepcidin expression by stimulating the transcription of matripase (MT-2) and furin, which cleaves HJV into its soluble form (56), in response to HIF (56). Inflammation impacts hepcidin production as hepcidin is an acute-phase protein whose expression is up-regulated in response to infection and inflammation caused by diseases or parasites. The underlying mechanism of this response is mediated by the inflammatory cytokine interleukin-6 (IL-6). IL-6 binds to its receptor and stimulates hepcidin synthesis via JAK2- STAT3 pathway in collaboration with BMP-SMAD signaling (42). Additionally, animal data show that hepcidin expression induction by inflammation can be potentiated by activin B via the activation of the BMP-SMAD signaling pathway (57). A brief schematic of hepatic hepcidin regulation signaling pathways is presented in Figure 3.1. c. Assessment of Iron Status Accurate assessment of iron status is important to not only predict iron demands but also to effectively intervene when needed to combat both iron deficiency and iron overload. The WHO recommends the use of multiple indicators to assess iron status and suggests 5 measures to characterize iron status: Hb, zinc protoporphyrin, mean cell volume, serum TfR, and serum ferritin (SF) (58). A brief description of iron status indicators and advantages/disadvantages are presented in Table 1.1. Measurements of SF and TfR are considered to provide the best approach to evaluate the iron status of populations when inflammation is absent (58). Serum ferritin is a measure of body iron stores and low concentrations (<12-15 ug/L) of SF indicate that iron stores are depleted. When inflammation is present, however, SF is not a useful indicator of 12 iron stores because concentrations of SF increase as a result of the acute phase response even if iron stores are low. Soluble TfR (sTfR) reflects the intensity of erythropoiesis and the cellular demands for iron. An increase in sTfR concentrations indicates increased cellular iron requirements or insufficient iron supply. Although sTfR appears to be less impacted by inflammation when compared SF, sTfR concentrations increase with general inflammation, increased erythropoiesis from malaria infection, or RBC disorders (59). In addition to use of individual measures of SF or sTfR, the ratio of sTfR to SF is often used to evaluate iron status. The logarithm of the ratio of sTfR to SF concentrations is linearly related to total body iron (TBI) when expressed as mg/kg body weight (Table 1.1). This calculated value provides a quantitative estimate of the size of the body iron stores; when iron is present values > 0 mg/kg are obtained and individuals that are iron deficient values < 0 mg/kg are obtained (60). Similarly, because both SF and sTfR are impacted by inflammation, TBI is also influenced by inflammation as this measure is obtained using both SF and sTfR. To control for limitations in serum ferritin as a biomarker when inflammation or infection is present, acute phase proteins including C-reactive protein (CRP) and α-1-antichymotrypsin can be measured to adjust for inflammatory conditions (58,61). Although not all anemia is caused by iron deficiency, Hb concentrations can provide useful information about the severity of iron deficiency. Hemoglobin concentrations are a key indicator of functional iron deficiency (ID). Hemoglobin can be used to diagnose anemia and as an initial screening indicator for ID , however, Hb concentrations are also impacted by smoking status, altitude, race, stage of gestation, thalassemia, vitamin A deficiency, and deficiencies of folic acid and vitamin B12 in certain settings (58). 13 Table 1.1: Biochemical iron status indicators Biochemical What this Biomarker Reflects Advantages Disadvantages Indicator Inexpensive; simple to Hemoglobin (Hb) Measure of anemia, reflect amount of Low sensitivity/specificity; measure; important health (g/dL) functional iron in body late indicator consequences Hematocrit (Ht) Proportional volume of RBCs in whole Simple to measure Same as Hb (%) blood Transferrin Indicator of tissue iron availability. Lacks standardization, affected receptor (TfR) Reflects balance between cellular iron Unaffected by inflammation by the rate of erythropoiesis (μg/L) requirements and iron supply Acute phase protein, affected Serum ferritin Reflects iron status; well Indicator of storage iron by inflammation and sub- (μg/L) standardized clinical infection - [log (serum transferrin Total Body Iron Measure of full range of iron receptor/serum ferritin) – Same as TfR and ferritin (TBI) (mg/kg) status 2.8229]/0.1207 Production diminished when Affected by inflammation and Hepcidin Regulator of iron homeostasis iron reserves depleted sub-clinical infection Varies diurnally and after Serum iron Iron bound to transferrin and ferritin Measure of iron supply to the meals; sample easily (ug/dL) (minor) in blood bone marrow and other tissues contaminated with iron from outside sources Total iron binding Large overlap between normal Total capacity of circulating transferrin Increased in iron deficiency; capacity (TIBC) values and values in iron bound to iron low in inflammatory disorders (μg/dL) deficiency Transferrin Proportion of transferrin Serum iron/TIBC Same as serum iron and TIBC saturation (%) bound to iron Zinc Useful in young children; Increased in iron deficiency, Indicates lack of iron to developing protoporphyrin whole blood or dried spots can inflammatory disorders, RBCs (μmol/mol) be assayed exposure to lead d. Stable Isotope Methodology to Measure Iron Absorption Given the unique role of iron absorption in iron homeostasis, a reliable and accurate method to measure iron absorption and utilization is needed. Stable iron isotope techniques are an invaluable and safe approach that can be used to measure oral iron absorption from different foods and to investigate factors impacting iron bioavailability. An isotope is an atom whose 14 nucleus contains the same number of protons but a different number of neutrons. Isotopes of the same element differ from each other in their atomic mass due to a variable number of neutrons. Iron isotopes can either be radioisotopes, which emit ionizing radiation, or stable isotopes which are non-radioactive, occurring naturally in fixed amounts in nature. Stable iron isotopes can be used as tracers, that are deliberately administered to humans to study the metabolism of iron. There are four naturally occurring stable iron isotopes and three of these (54Fe, 57Fe and 58Fe) are found at natural abundances under 6% (5.85%, 2.11% and 0.28% respectively). The erythrocyte iron incorporation method is the most commonly used method to measure iron absorption. The enrichment of orally administered stable iron isotopes in Hb is used as a proxy for iron absorption because the majority (80% or 90%) of absorbed iron is incorporated into RBCs in healthy adults (62) or infants and children (19,63,64) within 14 days. Thus, blood samples can be collected 14-days after administration of stable iron tracer and the enrichment of stable iron isotopes in whole blood can be measured to assess iron absorption. Thermal ionization mass spectrometry (TIMS) is often utilized to quantify the amount of iron isotope incorporated into RBCs after first digesting the blood, drying down the blood digests; isolating iron from the digests using anion exchange chromatography; loading the extracted iron onto rhenium filaments and measuring all four stable iron isotopes using TIMS. More details of the stable iron isotope method are presented in Chapter 3. II. Physiological influences on iron homeostasis Different life stages and corresponding physiological states have a significant impact on iron metabolism and homeostasis. The alteration of regular iron homeostasis occurs primarily in 15 individuals with increased iron requirements such as infants and children, adolescents during their growth spurt, women during their childbearing years, pregnant women, and in the elderly who may have pathophysiological changes in organ function and an increased inflammatory burden. The iron stores at birth are very important in maintaining adequate iron status over the first 6 months of life (65). The size of the iron stores at birth may be affected by birth weight, gestational age, timing of umbilical cord clamping and maternal iron status. Iron is transferred from maternal blood to the fetus with most fetal iron acquisition occurring during the third trimester. Early cord clamping and maternal IDA have a negative effect on iron status of newborn (65). Studies in human infants (66,67) and experimental animals suggest that the regulatory capacity of iron homeostasis is limited in young infants, mainly caused by a lack of regulation of DMT1 and ferroportin (68,69), and hyporesponsive to hepcidin (70). Assuming an infant is born with sufficient iron reserves, at about 4 months of age these iron stores are typically exhausted and the requirement for dietary iron increases to support the infant’s rapid growth. Rapid growth with high iron demands places infants and young children at particular risk for IDA, especially those aged 6–24 months (71). Iron deficiency and IDA can have a significant impact on infants’ and children’s development and health including altering immune status, increasing risk of morbidity, growth retardation, and adverse effects on cognition that may or may not be reversible with iron treatment (71). Iron is necessary for normal anatomic development of the fetal brain, myelination, and the development and function of the dopamine, serotonin, and norepinephrine systems, and iron also modifies the epigenetic landscape of the 16 brain (72). Adolescents are another at risk group for developing IDA due to their rapid growth and increased iron demands of puberty, especially among adolescent girls due to menstrual losses. Imbalances between nutritional iron intake and physiologic blood loss through menses or gestational iron losses are the primary reasons for ID and IDA in women in childbearing age (71,73). Menstrual blood losses may range from 25–50 mg of iron per cycle depending on the women’s Hb concentration and the individual pattern of menstruation, and losses can be substantially greater in women with menorrhagia (73). During pregnancy, iron requirements increase substantially to support fetoplacental development and maternal adaptations to pregnancy. In total, blood volume increases ~1.5 L to facilitate the blood flow in the uterus and placenta for nutrient and oxygen delivery to the fetus and to blunt the effects of blood loss at delivery (74). Insufficient increases in plasma volume have been associated with pathologies such as intrauterine growth restriction and preeclampsia (13,75). Iron requirements increase as pregnancy progresses. In the first trimester of pregnancy, the iron requirements (~0.8 mg/d) are lower than observed among non-pregnant women (1 - 1.5mg/day) (76) because of the cessation of menses. As pregnancy advances, maternal RBC mass increases and placental and fetal growth accelerates, physiologic iron requirements rise to 3.0 - 7.5 mg/d in the third trimester (13,73). To meet these increasing iron requirements, maternal hepcidin may be actively suppressed during pregnancy to increase both dietary iron absorption and mobilization of iron from stores (13). In healthy human pregnancies, maternal hepcidin concentrations decrease in the second and third trimesters, thereby facilitating an increased nonheme and heme iron absorption (77) as gestation progresses and to allow iron to be 17 mobilized from stores into the circulation (13). Other than hepcidin, EPO and sTfR concentrations increase as pregnancy progresses to support expanding maternal erythropoiesis (78–80). Serum ferritin concentrations gradually decrease reaching their lowest concentrations in the third trimester due to hemodilution and iron mobilization from stores to support the demands of the placenta and fetus (13). In addition, in an animal model of pregnancy, both DMT1 and the ferrireductase duodenal cytochrome B concentrations have been found to be increased (81). Although physiological iron requirements do not differ between adult and elderly men and post-menopausal and elderly women (82), there is growing evidence that iron metabolism is affected by the aging process. Hemoglobin concentrations have been reported to decline with advancing age, and this decline appears to increase after the age of 80, particularly in men (83). In addition, the prevalence of anemia increases with age within the elderly population. Anemia in the elderly may be caused by a number of physiological factors including reduced efficiency of iron absorption and chronic disease (84). Chronic low-grade inflammation in the elderly can lead to less efficient iron absorption under the regulation by acute phase protein, hepcidin (84). Moreover, moderate or severe kidney failure is one of the causes of anemia in elderly, due to compromised EPO production (85). Anemia in the elderly has adverse health implications including a decline in physical performance, cognitive impairment, increased susceptibility to falling, frailty, and mortality (86). On the other hand, there are potential adverse effects of elevated iron stores in middle-aged and older people. Elevated iron stores have been found to be associated with increased risk of chronic diseases such as heart disease, cancer and type 2 diabetes mellitus (84). 18 III. Impact of environmental factors on iron homeostasis There are many external environmental factors that impact iron status and iron absorption and can lead to disorders of iron balance. Lifestyle factors such as smoking, and alcohol consumption have been reported to be associated with iron homeostasis. Studies have exhibited that cigarette smoking induced significant dysregulation of iron homeostasis in the lung and lower sTfR values have been found in heavy smokers (1,87). Smoking can result in chronic hypoxic conditions and compromised lung capacity. The Hb distribution curve is shifted to the right in smokers and this has been found to be positively associated with the intensity of cigarette use (88). Other hematological indicators such as the hematocrit, MCV, MCH, MCHC, and RDW levels have also been found to be elevated in smokers to offset their compromised lung capacity (89). Alcohol consumption is another lifestyle factor influencing iron homeostasis. High concentrations of SF have been positively correlated with the degree of alcohol consumption, and even mild to moderate alcohol consumption has been shown to increase the prevalence of iron overload (1). These effects may be mediated by changes in hepcidin production as animal studies have demonstrated that alcohol can down-regulate hepcidin expression in the liver (90). Since there are no regulated biological excretion pathways to rid excess iron from the body, phlebotomy is used as an effective approach to treat iron overload. Blood donation can therefore have a negative impact on body iron stores. Results from the REDS-II Donor Iron Status Evaluation Study indicate that iron depletion is common in frequent blood donors, who were found to have significantly lower SF concentrations than non-blood donors or people who 19 had not donated blood for the past two years (91,92). Ferritin levels in blood donors were found to be inversely correlated with the cumulated number of blood donations, inter-donation intervals and donation intensity (91–94). Although only 1-2 mg/day of absorbed dietary iron is needed to offset the typical amounts of iron lost per day, long term dietary iron deficit would lead to ID and eventually result in IDA. At a population level, dietary iron absorption is generally considered to be one of the most important determinants of iron status (95). Dietary iron absorption not only depends on physiological requirements and existing body iron stores, but it is also influenced by the quantity, form of iron and other dietary components. Non-heme iron absorption exhibits significant variability between otherwise healthy individuals with values typically ranging from 2% to more than 20% (1,27). Iron absorption can be enhanced by dietary components including ascorbic acid, organic acids (malic, lactic and tartaric acids) and animal tissues (meat, poultry or fish) whereas it can be inhibited by calcium, phytate, tannins, other polyphenols, and soy proteins (96,97). Food components that are inhibitors of non-heme iron absorption generally bind iron in the gastrointestinal tract and prevent its absorption, while enhancers of iron absorption are food components that weaken or prevent iron binding to inhibitory compounds or work by reducing ferric iron to ferrous iron (1). Oral iron supplementation is a cost-effective strategy to treat ID and IDA in addition to improving dietary iron intake and bioavailability. Iron supplementation can be targeted to high- risk groups such as children and pregnant women who have higher physiological iron demands. Randomized control trials have reported that iron supplementation can improve Hb 20 concentrations in infants (98) and iron-deficient or anemic children (99) and can improve iron status (indicated by SF) in non-anemic iron-deficient adolescent girls (100). In addition, oral iron supplementation can reduce the risk of maternal anemia and iron deficiency in pregnancy (101). On the other hand, excessive iron supplementation of infants may lead to increased risk of infection, impaired growth and disturbed absorption or metabolism of other minerals (65,102,103) and may lead to dysbiosis in the gut microbiome (104). There has been controversy regarding the effects of exogenous iron administration in children in parasitic-endemic areas (16,17), particularly among children who are iron replete (105). Studies conducted in malaria- endemic areas have found that untargeted oral iron supplementation was associated with increased risk of contracting infectious diseases and increased mortality among preschool children (17,18). This effect, however, was not found in Nepal where there is no malaria, thus iron supplementation in malaria-endemic areas may be risky (106). Malaria is currently one of the most geographically widespread and life-threatening diseases caused by Plasmodium parasites that transmit malaria to people through the bites of infected female Anopheles mosquitoes. Malaria is responsible for the deaths of an estimated 0.6 million people per year (17). In 2019, there were an estimated 229 million cases of malaria worldwide, and children under 5 years of age were the most vulnerable group (107). Another common infection is soil-transmitted helminth infections, which can be caused by the main species of the roundworm (Ascaris lumbricoides), the whipworm (Trichuris trichiura) or by hookworms (Necator americanus and Ancylostoma duodenale). More than 1.5 billion people are infected with soil-transmitted helminth infections worldwide, with the greatest numbers 21 occurring in sub-Saharan Africa, the Americas, and East Asia. Over 267 million preschool-age children and over 568 million school-age children live in areas where these parasites are intensively transmitted (108). Children living in helminth- and malaria-endemic areas frequently suffer from chronic parasitemia and low-grade inflammation. These parasitic infections coupled with low dietary iron intakes increase the risk of IDA in many deprived areas. Mechanisms by which malaria and helminth infection impair iron homeostasis are multifactorial. Malaria may exacerbate ID by sequestrating iron into malarial pigment (hemazoin) (109) and by blocking iron recycling from senescent RBCs. It may also shorten the RBC lifespan due to erythrophagocytosis of both parasitized and non-parasitized RBC’s (110,111). Helminth infections can lead to chronic gastrointestinal blood loss from ingestion of RBCs by parasites or tissue trauma caused by parasites attachment and feeding, which increase the risk of ID and anemia (112). In addition, inflammation caused by malarial and helminth infections may further compromise iron absorption due to increased serum hepcidin concentrations by chronic inflammation (110,111). Because of the controversial effects of oral supplementation on children residing in parasites-endemic areas, the WHO recommends screening for ID before iron supplementation is initiated in these locations (113). In addition, WHO recommends provision of iron supplements to children, but this should be implemented in conjunction with measures to prevent, diagnose and treat malaria for optimal results on children’s health. Iron supplementation combined with anti-malarial and anti-helminth medication is commonly used as an effective strategy for managing anemia in malaria-endemic areas (114). 22 IV. Genetic influences on iron homeostasis Tight control of iron homeostasis is needed to prevent ID or iron overload but to date, only about 30% of the interindividual variability in iron absorption can be captured by or iron regulatory hormones or iron status biomarkers (20–24). Iron status as determined using SF, sTfR and Hb concentrations varies significantly by racial/ethnic groups and different geographic areas (115–119). Ethnic differences in iron status indicators have been reported in large epidemiological studies or large cohort studies over the past few decades. For example, the largest epidemiological study to date that evaluated iron stores as a function of ethnicity was the HEIRS (Hemochromatosis and Iron Overload Screening) study. This study recruited 101,168 primary care adults aged 25 years or older from the United States and Canada, and evaluated SF and TSAT as a function of ethnicity (self-reported as Hispanic, European, African American, Asian, Pacific Islander, Native American) (120). These epidemiological studies have found that Asians have higher adjusted means of serum ferritin and transferrin saturation compared to any other population group studied, even after excluding polymorphisms previously associated with iron overload (115,118) and African Americans have lower mean Hb concentrations and higher SF concentrations compared to Caucasians (121,122). Race-ethnic differences in the risk of ID or iron overload have also been reported as highlighted by the CDC cut-offs used to denote anemia (123). Findings from the O’Brien and Gu laboratories are in agreement with the epidemiologic data and have observed population-specific differences in iron absorption between East Asians and Europeans, and the differences noted persisted even after controlling for iron status (20). 23 These differences might indicate that the storage threshold at which East Asians down-regulate iron absorption is higher. Increased iron absorption and its impact on augmentation of body iron stores across age groups suggest that East Asian populations may be at greater risk of iron overload related conditions at maturity, which might explain East Asian populations having greater risk of diabetes at a lower BMI (124,125). Diet might be a plausible reason for the ethnic disparities in iron status because studies have found that there are associations between dietary constituents (such as red meat and vitamin C) and iron status (116,117). However, a randomized clinical trial conducted in elderly Europeans showed that changing from a Western to a Mediterranean-style diet for one year had no overall effect on iron status (116). On the other hand, genetic variations might underlie the observed ethnic differences in iron status. The HFE gene was first identified as the causative gene in hereditary hemochromatosis (HH) but the C282Y and H63D mutations of HFE that manifest at very high frequencies in Northern Europeans, are nearly absent in Asian and Pacific Island populations (119,126,127). In addition, a unique East Asian-prevalent HFE haplotype has been identified and may be a result of adaptive evolution (128). Clearly, there are unidentified genetic contributions that underlie the observed ethnic differences in iron status. Multiple genome-wide association studies (GWAS) and candidate gene association studies have been undertaken to investigate the genetic contributions to variations in iron status between and within populations. These studies have identified pathogenic mutations in some iron-related genes that associated with ID or iron overload and identified polymorphisms within iron- and even non-iron related genes that associated with variations in iron traits among healthy populations. 24 V. Significance and Summary This dissertation project investigates three major factors that influence iron homeostasis: physiology, environment, and genetics. These three factors will be addressed in three specific aims of this dissertation that will be individually detailed and discussed in Chapters 2 through 4 (Figure 1.2). Figure 1.2 Three major factors that influence iron homeostasis and the corresponding dissertation chapters addressing these factors. For physiological factors, pregnancy will be focused on in this research. Maternal anemia has been reported to be associated with poor birth outcomes such as increased risks of low birth weight (LBW), SGA, preterm birth, stillbirth, perinatal and neonatal mortality (14,129) as well as adverse maternal outcomes including postpartum hemorrhage, preeclampsia, blood transfusion and maternal mortality (14,130). However, the policy of providing universal iron supplementation to all women in developed countries remains controversial. The US Preventive Services Task Force (USPSTF) determined that the current evidence was insufficient to recommend routine iron supplementation of pregnant women to prevent adverse maternal health 25 and birth outcomes (131). Limited normative data are available for maternal Hb in pregnancy and data to estimate the current prevalence of IDA among healthy US pregnant women. Therefore, Aim 1 (in Chapter 2) is designed to characterize the distribution of maternal Hb and the prevalence of anemia in a large cohort of pregnant women and to identify factors associated with risk of anemia in this cohort. With respect to environmental factors and their impact on iron status, to date little is known about iron absorption and utilization in children with mild chronic parasitic infections (19). In addition, how oral iron supplementation impacts iron utilization and long-term iron metabolism in children residing in parasite-endemic areas is unknown. Therefore, Aim 2 (in Chapter 3) addresses the impact of malarial and helminth infections on iron incorporation into RBCs and identifies variables associated with the change in RBC iron enrichment over an 84-day period post dosing in a group of toddlers at high risk for malarial and helminth infections. Lastly, understanding the impact of genetic polymorphisms and evolutionary origins on iron homeostasis may help guide dietary recommendations to minimize the risks of chronic diseases and guide future genome-informed nutritional practices. However, no published studies to date have summarized data on genetic variants that are associated with iron metabolism and highlighted how these genetic variants may differ across ethnic groups. Multiple published reports of iron status exist but few attempts have been made to compile these existing data to summarize current findings and knowledge on ethnic differences in iron status. Therefore, to fill this knowledge gap, Aim 3 (in Chapter 4) uses existing published data to statistically explore 26 differences in iron status among various ethnic/racial populations and to summarize and review published data on genetic determinants of iron status. 27 REFERENCES 1. Lynch S, Pfeiffer CM, Georgieff MK, et al. Biomarkers of Nutrition for Development (BOND)-Iron review. J Nutr. 2018;148(suppl_1):1001S-1067S. 2. Miller JL. Iron deficiency anemia: a common and curable disease. Cold Spring Harb Perspect Med. 2013;3(7):. 3. Camaschella C. Iron deficiency. Blood. 2019;133(1):30–39. 4. Montonen J, Boeing H, Steffen A, et al. Body iron stores and risk of type 2 diabetes: results from the European Prospective Investigation into Cancer and Nutrition (EPIC)- Potsdam study. Diabetologia. 2012;55(10):2613–2621. 5. Gan W, Guan Y, Wu Q, et al. Association of TMPRSS6 polymorphisms with ferritin, hemoglobin, and type 2 diabetes risk in a Chinese Han population. Am J Clin Nutr. 2012;95(3):626–632. 6. Gujja P, Rosing DR, Tripodi DJ, Shizukuda Y. Iron overload cardiomyopathy: better understanding of an increasing disorder. J Am Coll Cardiol. 2010;56(13):1001–12. 7. Eftekhari MH, Mozaffari-Khosravi H, Shidfar F, Zamani A. Relation between Body Iron Status and Cardiovascular Risk Factors in Patients with Cardiovascular Disease. Int J Prev Med. 2013;4(8):911–6. 8. Basuli D, Stevens RG, Torti FM, Torti S V. Epidemiological associations between iron and cardiovascular disease and diabetes. Front Pharmacol. 2014;5:117. 28 9. Torti S V., Torti FM. Iron and cancer: more ore to be mined. Nat Rev Cancer. 2013;13(5):342–355. 10. Thévenod F. 15. Iron and Its Role in Cancer Defense: A Double-Edged Sword. Met Dev Action Anticancer Agents. 2018;437–468. 11. Daugherty AM, Raz N. Appraising the Role of Iron in Brain Aging and Cognition: Promises and Limitations of MRI Methods. Neuropsychol Rev. 2015;25(3):272–87. 12. Ward RJ, Dexter DT, Crichton RR. Neurodegenerative diseases and therapeutic strategies using iron chelators. J Trace Elem Med Biol. 2015;31:267–273. 13. Fisher AL, Nemeth E. Iron homeostasis during pregnancy. Am J Clin Nutr. 2017;106:1567S-1574S. 14. Young MF, Oaks BM, Tandon S, et al. Maternal hemoglobin concentrations across pregnancy and maternal and child health: a systematic review and meta-analysis. Ann N Y Acad Sci. 2019;1450(1):47. 15. Ganz T, Nemeth E. Iron homeostasis in host defence and inflammation. Nat Publ Gr. 2015; 16. Barffour MA, Schulze KJ, Coles CL, et al. High Iron Stores in the Low Malaria Season Increase Malaria Risk in the High Transmission Season in a Prospective Cohort of Rural Zambian Children. J Nutr. 2017; 29 17. Spottiswoode N, Duffy PE, Drakesmith H. Iron, anemia and hepcidin in malaria. Front Pharmacol. 2014; 18. Sazawal S, Black RE, Ramsan M, et al. Effects of routine prophylactic supplementation with iron and folic acid on admission to hospital and mortality in preschool children in a high malaria transmission setting: Community-based, randomised, placebo-controlled trial. Lancet. 2006; 19. Doherty CP, Cox SE, Fulford AJ, et al. Iron incorporation and post-malaria anaemia. PLoS One. 2008; 20. Ye K, Cao C, Lin X, O’Brien KO, Gu Z. Natural selection on HFE in Asian populations contributes to enhanced non-heme iron absorption. BMC Genet. 2015;16(1):61. 21. Young MF, Glahn RP, Ariza-Nieto M, et al. Serum hepcidin is significantly associated with iron absorption from food and supplemental sources in healthy young women. Am J Clin Nutr. 2009;89(2):533–538. 22. Roe MA, Heath ALM, Oyston SL, et al. Iron absorption in male C282Y heterozygotes. Am J Clin Nutr. 2005;81(4):814–821. 23. Zimmermann MB, Troesch B, Biebinger R, et al. Plasma hepcidin is a modest predictor of dietary iron bioavailability in humans, whereas oral iron loading, measured by stable- isotope appearance curves, increases plasma hepcidin. 2009;90(5):1280–1287. 24. Roe MA, Collings R, Dainty JR, Swinkels DW, Fairweather-Tait SJ. Plasma hepcidin 30 concentrations significantly predict interindividual variation in iron absorption in healthy men. Am J Clin Nutr. 2009;89(4):1088–1091. 25. Dev S, Babitt JL. Overview of iron metabolism in health and disease. Hemodial Int. 2017;21 Suppl 1(Suppl 1):S6–S20. 26. Knutson MD. Iron transport proteins: Gateways of cellular and systemic iron homeostasis. J Biol Chem. 2017;292(31):12735–12743. 27. Abbaspour N, Hurrell R, Kelishadi R. Review on iron and its importance for human health. J Res Med Sci. 2014;19(2):164–74. 28. Andrews NC. Iron homeostasis: insights from genetics and animal models. Nat Rev Genet. 2000;1(3):208–217. 29. Ganz T, Nemeth E. Regulation of iron acquisition and iron distribution in mammals. Biochim Biophys Acta - Mol Cell Res. 2006;1763(7):690–699. 30. Fleming RE, Ponka P. Iron Overload in Human Disease. N Engl J Med. 2012;366(4):348– 359. 31. Anderson GJ, Frazer DM. Current understanding of iron homeostasis. Oxford University Press; 2017. 32. Jenkitkasemwong S, Wang C-Y, Coffey R, et al. SLC39A14 Is Required for the Development of Hepatocellular Iron Overload in Murine Models of Hereditary 31 Hemochromatosis. Cell Metab. 2015;22(1):138–150. 33. Chen M, Cabantchik ZI, Chan S, Chan GC, Cheung Y. Iron overload and apoptosis of HL-1 cardiomyocytes: effects of calcium channel blockade. PLoS One. 2014;9(11):e112915. 34. Knutson MD. Non-transferrin-bound iron transporters. Free Radic Biol Med. 2019;133:101–111. 35. Andrews NC. Disorders of Iron Metabolism. N Engl J Med. 1999;341(26):1986–1995. 36. Camaschella C, Nai A, Silvestri L. Iron metabolism and iron disorders revisited in the hepcidin era. Haematologica. 2020;105(2):260–272. 37. Jelkmann W. Erythropoietin after a century of research: younger than ever. Eur J Haematol. 2007;78(3):183–205. 38. Gupta N, Wish JB. Hypoxia-Inducible Factor Prolyl Hydroxylase Inhibitors: A Potential New Treatment for Anemia in Patients With CKD. Am J Kidney Dis. 2017;69(6):815– 826. 39. Krause A, Neitz S, Mägert HJ, et al. LEAP-1, a novel highly disulfide-bonded human peptide, exhibits antimicrobial activity. FEBS Lett. 2000;480(2–3):147–50. 40. Aschemeyer S, Qiao B, Stefanova D, et al. Structure-function analysis of ferroportin defines the binding site and an alternative mechanism of action of hepcidin. Blood. 32 2018;131(8):899–910. 41. Goh JB, Wallace DF, Hong W, Subramaniam VN. Endofin, a novel BMP-SMAD regulator of the iron-regulatory hormone, hepcidin. Sci Rep. 2015;5(1):1–12. 42. Silvestri L, Nai A, Dulja A, Pagani A. Hepcidin and the BMP-SMAD pathway: An unexpected liaison. Vitam Horm. 2019;110:71–99. 43. Wang RN, Green J, Wang Z, et al. Bone Morphogenetic Protein (BMP) signaling in development and human diseases. Genes Dis. 2014;1(1):87–105. 44. Muckenthaler MU, Rivella S, Hentze MW, Galy B. A Red Carpet for Iron Metabolism. Cell. 2017; 45. Core AB, Canali S, Babitt JL. Hemojuvelin and bone morphogenetic protein (BMP) signaling in iron homeostasis. Front Pharmacol. 2014;5 MAY:104. 46. Wu XG, Wang Y, Wu Q, et al. HFE interacts with the BMP type I receptor ALK3 to regulate hepcidin expression. Blood. 2014; 47. Zhao N, Nizzi CP, Anderson SA, et al. Low intracellular iron increases the stability of Matriptase-2. J Biol Chem. 2015; 48. Colucci S, Pagani A, Pettinato M, et al. The immunophilin FKBP12 inhibits hepcidin expression by binding the BMP type I receptor ALK2 in hepatocytes. Blood. 2017;130(19):2111–2120. 33 49. Kautz L, Jung G, Valore E V, et al. Identification of erythroferrone as an erythroid regulator of iron metabolism. Nat Genet. 2014;46(7):678–684. 50. Robach P, Gammella E, Recalcati S, et al. Induction of erythroferrone in healthy humans by micro-dose recombinant erythropoietin or high-altitude exposure. Haematologica. 2020;haematol.2019.233874. 51. Ganz T. Erythropoietic regulators of iron metabolism. Free Radic Biol Med. 2019;133:69– 74. 52. Arezes J, Foy N, McHugh K, et al. Erythroferrone inhibits the induction of hepcidin by BMP6. Blood. 2018;132(14):1473–1477. 53. Aschemeyer S, Gabayan V, Ganz T, Nemeth E, Kautz L. Erythroferrone and matriptase-2 independently regulate hepcidin expression. Am J Hematol. 2017;92(5):E61–E63. 54. Zhang AS, Enns CA. A long sought after “receptor” for ERFE? Blood. 2018;132(14):1463–1464. 55. Wang C-Y, Xu Y, Traeger L, et al. Erythroferrone lowers hepcidin by sequestering BMP2/6 heterodimer from binding to the BMP type I receptor ALK3. Blood. 2020;135(6):453–456. 56. Silvestri L, Pagani A, Camaschella C. Furin-mediated release of soluble hemojuvelin: A new link between hypoxia and iron homeostasis. Blood. 2008;111(2):924–931. 34 57. Canali S, Core AB, Zumbrennen-Bullough KB, et al. Activin B Induces Noncanonical SMAD1/5/8 Signaling via BMP Type I Receptors in Hepatocytes: Evidence for a Role in Hepcidin Induction by Inflammation in Male Mice. Endocrinology. 2016;157(3):1146– 1162. 58. WHO, World Health Organization, WHO. WHO | Assessing the iron status of populations. World Health Organization; 2012. 59. Suchdev PS, Williams AM, Mei Z, et al. Assessment of iron status in settings of inflammation: Challenges and potential approaches. Am J Clin Nutr. 2017;106(Suppl 6):1626S-1633S. 60. Pfeiffer CM, Looker AC. Laboratory methodologies for indicators of iron status: strengths, limitations, and analytical challenges. Am J Clin Nutr. 2017;106(Supplement 6):1606S-1614S. 61. Namaste SM, Aaron GJ, Varadhan R, Peerson JM, Suchdev PS. Methodologic approach for the Biomarkers Reflecting Inflammation and Nutritional Determinants of Anemia (BRINDA) project. Am J Clin Nutr. 2017;106(Suppl 1):333S-347S. 62. Assessment of Iron Bioavailability in Humans Using Stable Iron Isotope Techniques. Vienna: INTERNATIONAL ATOMIC ENERGY AGENCY; 2012. 63. Chen Z, Griffin IJ, Plumlee LM, Abrams SA. High resolution inductively coupled plasma mass spectrometry allows rapid assessment of iron absorption in infants and children. J 35 Nutr. 2005; 64. Abrams SA. Using stable isotopes to assess mineral absorption and utilization by children. Am J Clin Nutr. 1999;70(6):955–964. 65. Lönnerdal B. Development of iron homeostasis in infants and young children. Am J Clin Nutr. 2017;106:1575S-1580S. 66. Domellöf M, Lönnerdal B, Abrams SA, Hernell O. Iron absorption in breast-fed infants: Effects of age, iron status, iron supplements, and complementary foods. Am J Clin Nutr. 2002;76(1):198–204. 67. Domellöf M, Cohen RJ, Dewey KG, et al. Iron supplementation of breast-fed Honduran and Swedish infants from 4 to 9 months of age. J Pediatr. 2001;138(5):679–687. 68. Leong WI, Bowlus CL, Tallkvist J, Lönnerdal B. Iron supplementation during infancy - Effects on expression of iron transporters, iron absorption, and iron utilization in rat pups. Am J Clin Nutr. 2003;78(6):1203–1211. 69. Leong WI, Bowlus CL, Tallkvist J, Lönnerdal B. DMT1 and FPN1 expression during infancy: Developmental regulation of iron absorption. Am J Physiol - Gastrointest Liver Physiol. 2003;285(6 48-6):. 70. Darshan D, Wilkins SJ, Frazer DM, Anderson GJ. Reduced expression of ferroportin-1 mediates hyporesponsiveness of suckling rats to stimuli that reduce iron absorption. Gastroenterology. 2011;141(1):300–309. 36 71. Hermoso M, Vucic V, Vollhardt C, et al. The Effect of Iron on Cognitive Development and Function in Infants, Children and Adolescents: A Systematic Review. Ann Nutr Metab. 2011;59(2–4):154–165. 72. Cusick SE, Georgieff MK. The Role of Nutrition in Brain Development: The Golden Opportunity of the “First 1000 Days.” J Pediatr. 2016;175:16–21. 73. Means RT. Iron Deficiency and Iron Deficiency Anemia: Implications and Impact in Pregnancy, Fetal Development, and Early Childhood Parameters. Nutrients. 2020;12(2):447. 74. Ramsay M. Normal hematological changes during pregnancy and the puerperium. Obstet Hematol Man. 2011;3–12. 75. de Haas S, Ghossein-Doha C, van Kuijk SMJ, van Drongelen J, Spaanderman MEA. Physiological adaptation of maternal plasma volume during pregnancy: a systematic review and meta-analysis. Ultrasound Obstet Gynecol. 2017;49(2):177–187. 76. Bothwell TH. Iron requirements in pregnancy and strategies to meet them. Am J Clin Nutr. 2000;72(1 SUPPL.): 77. Young MF, Griffin I, Pressman E, et al. Maternal hepcidin is associated with placental transfer of iron derived from dietary heme and nonheme sources. J Nutr. 2012;142(1):33– 39. 78. McMullin MF, White R, Lappin T, Reeves J, MacKenzie G. Haemoglobin during 37 pregnancy: Relationship to erythropoietin and haematinic status. Eur J Haematol. 2003;71(1):44–50. 79. Harstad TW, Mason RA, Cox SM. Serum Erythropoietin Quantitation in Pregnancy Using an Enzyme-Linked Immunoassay. Am J Perinatol. 1992;9(4):233–235. 80. Milman N, Agger AO, Nielsen OJ. Iron status markers and serum erythropoietin in 120 mothers and newborn infants: Effect of iron supplementation in normal pregnancy. Acta Obstet Gynecol Scand. 1994;73(3):200–204. 81. Millard KN, Frazer DM, Wilkins SJ, Anderson GJ. Changes in the expression of intestinal iron transport and hepatic regulatory molecules explain the enhanced iron absorption associated with pregnancy in the rat. Gut. 2004;53(5):655–660. 82. NIH. Iron - Health Professional Fact Sheet. US Dep Heal Hum Serv. 2018;18. 83. Nilsson-Ehle H, Jagenburg R, Landahl S, Svanborg A. Blood haemoglobin declines in the elderly: Implications for reference intervals from age 70 to 88. Eur J Haematol. 2000;65(5):297–305. 84. Fairweather-Tait SJ, Wawer AA, Gillings R, Jennings A, Myint PK. Iron status in the elderly. Mech Ageing Dev. 2014;136–137:22–28. 85. Wawer AA, Jennings A, Fairweather-Tait SJ. Iron status in the elderly: A review of recent evidence. Mech Ageing Dev. 2018;175:55–73. 38 86. Price EA, Mehra R, Holmes TH, Schrier SL. Anemia in older persons: Etiology and evaluation. Blood Cells, Mol Dis. 2011;46(2):159–165. 87. Zhang WZ, Butler JJ, Cloonan SM. Smoking-induced iron dysregulation in the lung. Free Radic Biol Med. 2019;133:238–247. 88. Nordenberg D, Yip R, Binkin NJ. The Effect of Cigarette Smoking on Hemoglobin Levels and Anemia Screening. JAMA J Am Med Assoc. 1990;264(12):1556–1559. 89. Elisia I, Lam V, Cho B, et al. The effect of smoking on chronic inflammation, immune function and blood cell composition. Sci Rep. 2020;10(1):. 90. Harrison-Findik DD. Role of alcohol in the regulation of iron metabolism. World J Gastroenterol. 2007;13(37):4925–4930. 91. Cable RG, Glynn SA, Kiss JE, et al. Iron deficiency in blood donors: Analysis of enrollment data from the REDS-II Donor Iron Status Evaluation (RISE) study. Transfusion. 2011;51(3):511–522. 92. Cable RG, Glynn SA, Kiss JE, et al. Iron deficiency in blood donors: The REDS-II Donor Iron Status Evaluation (RISE) study. Transfusion. 2012;52(4):702–711. 93. Milman N, Byg KE, Ovesen L. Iron status in Danes 1994: II: Prevalence of iron deficiency and iron overload in 1319 Danish women aged 40-70 years. Influence of blood donation, alcohol intake and iron supplementation. Ann Hematol. 2000;79(11):612–621. 39 94. Spencer BR, Guo Y, Cable RG, et al. Iron status and risk factors for iron depletion in a racially/ethnically diverse blood donor population. Transfusion. 2019;59(10):3146–3156. 95. Dainty JR, Berry R, Lynch SR, Harvey LJ, Fairweather-Tait SJ. Estimation of Dietary Iron Bioavailability from Food Iron Intake and Iron Status. PLoS One. 2014;9(10):e111824. 96. Hurrell R, Egli I. Iron bioavailability and dietary reference values. Am J Clin Nutr. 2010;91(5):1461S-1467S. 97. Hallberg L, Hulthén L. Prediction of dietary iron absorption: an algorithm for calculating absorption and bioavailability of dietary iron. Am J Clin Nutr. 2000;71(5):1147–1160. 98. Friel JK, Aziz K, Andrews WL, et al. A double-masked, randomized control trial of iron supplementation in early infancy in healthy term breast-fed infants. J Pediatr. 2003;143(5):582–586. 99. Iannotti LL, Tielsch JM, Black MM, Black RE. Iron supplementation in early childhood: health benefits and risks. Am J Clin Nutr. 2006;84(6):1261–1276. 100. Bruner AB, Joffe A, Duggan AK, Casella JF, Brandt J. Randomised study of cognitive effects of iron supplementation in non-anaemic iron-deficient adolescent girls. Lancet. 1996;348(9033):992–996. 101. Peña-Rosas JP, De-Regil LM, Garcia-Casal MN, Dowswell T. Daily oral iron supplementation during pregnancy. Cochrane Database Syst Rev. 2015;2015(7):1–527. 40 102. Lönnerdal B, Georgieff MK, Hernell O. Developmental physiology of iron absorption, homeostasis, and metabolism in the healthy term infant. J Pediatr. 2015;167(4):S8–S14. 103. Lönnerdal B. Excess iron intake as a factor in growth, infections, and development of infants and young children. Am J Clin Nutr. 2017;106:1681S-1687S. 104. Paganini D, Zimmermann MB. The effects of iron fortification and supplementation on the gut microbiome and diarrhea in infants and children: A review. Am J Clin Nutr. 2017;106(Suppl 6):1688S-1693S. 105. Neuberger A, Okebe J, Yahav D, Paul M. Oral iron supplements for children in malaria- endemic areas. Cochrane Database Syst Rev. 2016; 106. Georgieff MK, Krebs NF, Cusick SE. The Benefits and Risks of Iron Supplementation in Pregnancy and Childhood. Annu Rev Nutr. 2019;39:121–146. 107. World Health Organization. Malaria. WHO. 2021; 108. World Health Organization. Soil-transmitted helminth infections. WHO. 2020;391(10117):252–265. 109. Olivier M, Van Den Ham K, Shio MT, Kassa FA, Fougeray S. Malarial pigment hemozoin and the innate inflammatory response. Front Immunol. 2014; 110. Cherayil BJ. Pathophysiology of iron homeostasis during inflammatory states. J Pediatr. 2015; 41 111. Friedman JF, Kanzaria HK, McGarvey ST. Human schistosomiasis and anemia: The relationship and potential mechanisms. Trends Parasitol. 2005; 112. Crompton DWT, Nesheim MC. Nutritional impact of intestinal helminthiasis during the human life cycle. Annu Rev Nutr. 2002;22:35–59. 113. Cusick SE, Georgieff MK, Rao R. Approaches for reducing the risk of early-life iron deficiency-induced brain dysfunction in children. Nutrients. 2018; 114. Cusick SE, Opoka RO, Ssemata AS, et al. Delayed iron improves iron status without altering malaria risk in severe malarial anemia. Am J Clin Nutr. 2020;111(5):1059–1067. 115. Gordeuk VR, Brannon PM. Ethnic and genetic factors of iron status in women of reproductive age. Am J Clin Nutr. 2017;106:1594S-1599S. 116. Jennings A, Tang J, Gillings R, et al. Changing from a Western to a Mediterranean-style diet does not affect iron or selenium status: results of the New Dietary Strategies Addressing the Specific Needs of the Elderly Population for Healthy Aging in Europe (NU-AGE) 1-year randomized clinical trial in elderly Europeans. Am J Clin Nutr. 2019; 117. Hu PJ, Ley SH, Bhupathiraju SN, Li Y, Wang DD. Associations of dietary, lifestyle, and sociodemographic factors with iron status in Chinese adults: A cross-sectional study in the China Health and Nutrition Survey. Am J Clin Nutr. 2017;105(2):503–512. 118. Harris EL, McLaren CE, Reboussin DM, et al. Serum Ferritin and Transferrin Saturation in Asians and Pacific Islanders. Arch Intern Med. 2007;167(7):722. 42 119. Adams PC, Reboussin DM, Barton JC, et al. Hemochromatosis and iron-overload screening in a racially diverse population. N Engl J Med. 2005; 120. McLaren CE, Barton JC, Adams PC, et al. Hemochromatosis and iron overload screening (HEIRS) Study design for an evaluation of 100,000 primary care-based adults. Am J Med Sci. 2003;325(2):53–62. 121. Zakai NA, McClure LA, Prineas R, et al. Correlates of anemia in American blacks and whites: the REGARDS Renal Ancillary Study. Am J Epidemiol. 2009;169(3):355–64. 122. WHO | Assessing the iron status of populations. WHO. 2014; 123. CDC, Satcher D, Hughes JM, et al. Recommendations to Prevent and Control Iron Deficiency in the United States. MMWR Recomm Rep. 1998;12(47(RR–3)):210. 124. Hsu WC, Araneta MRG, Kanaya AM, Chiang JL, Fujimoto W. BMI cut points to identify at-risk Asian Americans for type 2 diabetes screening. Diabetes Care. 2015;38(1):150–8. 125. WHO Expert Consultation. Appropriate body-mass index for Asian populations and its implications for policy and intervention strategies. Lancet. 2004;363(9403):157–163. 126. Chang JG, Liu TC, Lin SF. Rapid diagnosis of the HLA-H gene Cys 282 Tyr mutation in hemochromatosis by polymerase chain reaction--a very rare mutation in the Chinese population. Blood. 1997;89(9):3492–3. 127. Sohda T, Yanai J, Soejima H, Tamura K. Frequencies in the Japanese Population of HFE 43 Gene Mutations. Biochem Genet. 1999;37(1/2):63–68. 128. Ye K, Gu Z. Recent Advances in Understanding the Role of Nutrition in Human Genome Evolution. Adv Nutr. 2011;2(6):486–496. 129. Rahman MM, Abe SK, Rahman MS, et al. Maternal anemia and risk of adverse birth and health outcomes in low- and middle-income countries: Systematic review and meta- analysis. Am J Clin Nutr. 2016;103(2):495–504. 130. Daru J, Zamora J, Fernández-Félix BM, et al. Risk of maternal mortality in women with severe anaemia during pregnancy and post partum: a multilevel analysis. Lancet Glob Heal. 2018;6(5):e548–e554. 131. Siu AL, Albert L. Siu. Screening for Iron Deficiency Anemia and Iron Supplementation in Pregnant Women to Improve Maternal Health and Birth Outcomes: U.S. Preventive Services Task Force Recommendation Statement. American College of Physicians; 2015. 44 CHAPTER 2 LONGITUDINAL CHANGES IN HEMOGLOBIN AND DETERMINANTS OF ANEMIA IN HEALTHY U.S. PREGNANT WOMEN* *Kang W, Irvine C, Pressman EK, O’Brien KO. Longitudinal changes in hemoglobin and determinants of anemia in healthy U.S. pregnant women. To be Submitted, The American Journal of Clinical Nutrition. 45 Abstract Background: Maternal anemia has been associated with poor birth outcomes. At present longitudinal data on hemoglobin concentrations across gestation are lacking and few data are available to estimate the current prevalence of anemia among U.S. pregnant women. Objective: This study aimed to characterize normative maternal hemoglobin (Hb) distributions, the prevalence of anemia and determinants of anemia in a large cohort of healthy U.S. pregnant women receiving prenatal care at an urban medical center. Methods: A 10-y retrospective medical chart review was undertaken of all healthy pregnancies that occurred between 2011 to 2020 at Strong Memorial Hospital and Highland Hospital in Rochester, NY. Data on Hb, serum ferritin and health and demographic information were abstracted from medical records. Mixed-effects logistic regression models were utilized to identify significant determinants of maternal anemia. Results: A total of 54,453 pregnancies were identified, of these 41,226 were uncomplicated pregnancies that met all study inclusion criteria. Mean Hb concentration was lowest in the second trimester and the prevalence of anemia increased in each trimester of pregnancy (P < 0.0001). The overall prevalence of anemia was 16.7%, which significantly increased to 25.1% by late gestation. Risk factors for anemia included Black race or Hispanic ethnicity, being underweight at entry into pregnancy, being multiparous or carrying a multiple birth pregnancy. The current CDC anemia cutoffs which reflected the 5th percentile of the reference population represented the 5th, 16th and the 27th percentiles of our population in each trimester. Conclusions: Anemia was evident in nearly 20% of otherwise healthy U.S. pregnant women and 46 the prevalence of anemia increased significantly as pregnancy progressed. Risk of maternal anemia was significantly higher in women that self-identified as Black race, or that were underweight, multiparous or carrying a multiple birth pregnancy. Current Hb cut-offs used to define anemia may be too low and may overestimate the prevalence of anemia. 47 Introduction Maternal anemia is a significant public health problem that is estimated to have an global prevalence of 41.8% affecting nearly 56.4 million pregnancies (1). Anemia during pregnancy has been associated with poor birth outcomes including increased risks of low birth weight (LBW), small-for-gestational age (SGA), preterm birth, stillbirth, perinatal and neonatal mortality (2,3) as well as adverse maternal outcomes including postpartum hemorrhage, preeclampsia, blood transfusion and maternal mortality (2,4). In the United States (U.S.), the 1999-2010 National Health and Nutrition Examination Survey (NHANES) reported an overall prevalence of gestational anemia of 2.6% (n = 1,283 pregnant women) (5), and based on the 2003-2012 NHANES data (n = 776 pregnant women) this prevalence increased to 8.8% (6). The trimester specific prevalence of anemia was reported using the NHANES 1999-2006 data which found that the risk of anemia increased nearly 5-fold as pregnancy progressed from 2.7% (n = 189), 2.2% (n = 416) and 10.8% (n = 384), in the first (T1), second (T2) and third (T3) trimester respectively (7). Of concern, the absolute number of pregnant women sampled in NHANES surveys has been relatively low, limiting the conclusions that can be generated from these cross-sectional data. Iron supplementation is commonly used as an effective treatment to combat anemia among pregnant women and a review of this practice by the US Preventive Services Task Force (USPSTF) in 2015 concluded routine iron supplementation during pregnancy did improve intermediate maternal hematologic indexes, such as serum ferritin (SF) and hemoglobin (Hb). However this USPSTF review concluded that the existing evidence was insufficient to recommend routine iron 48 supplementation to pregnant women to prevent adverse maternal health and birth outcomes (9). The NIH Office of Dietary Supplements followed up on this report with a workshop highlighting evidence gaps and research needs on iron screening and supplementation in iron-replete pregnant women. One of the key knowledge gaps identified was limited cross-sectional data, and lack of longitudinal data on the prevalence of iron deficiency (ID) and iron deficiency anemia (IDA) among U.S. pregnant women (9). The existing U.S. Centers for Disease Control and Prevention (CDC) cut-offs for gestational anemia were derived from 5th centile values obtained from longitudinal Hb data that was pooled from four small European studies undertaken more than 4 decades ago in healthy iron-supplemented pregnant women (n = 394) (10–13). The characteristics of current US pregnant women are dramatically different from the CDC reference populations in terms of racial composition, pre-pregnancy BMI (ppBMI) and supplemental iron intake (14). To address existing gaps in this literature, we undertook a 10-year retrospective medical chart review of healthy pregnant women receiving prenatal care at a large academic medical center. Normative data on hemoglobin concentrations were abstracted from medical records to provide normative data on maternal Hb concentrations across gestation and to identify factors associated with risk of maternal anemia. 49 Methods Study population and design A retrospective medical chart review was undertaken using demographic and clinical data extracted from medical records of all pregnant women receiving prenatal care from 2011 to 2020 at Strong Memorial Hospital and Highland Hospital in Rochester, NY. Data were extracted from medical records by an honest broker and were provided in deidentified form to study investigators. This study was approved by the Institutional Review Boards of the University of Rochester and Cornell University. Information on maternal demographic, anthropometric, and health-related characteristics were abstracted from the medical charts. Demographic information obtained included maternal age at delivery and self-reported race and ethnicity. Anthropometric data included maternal height measured during T1, self-reported pre-pregnancy body weight, and maternal weight across trimesters. BMI was calculated as weight [kg] / height2 [m2]. Women were classified into one of four BMI categories according to the CDC guidelines based on self-reported ppBMI as: underweight (< 18.5 kg/m2); normal weight (18.5 to <25.0 kg/m2); overweight (25.0 to <30.0 kg/m2); obese, (30.0 to < 40 kg/m2) or severely obese (≥ 40.0 kg/m2) (17). Gestational weight gain (GWG) in kg was calculated as weight at the latest T3 measure obtained minus the pre- pregnancy weight. Health-related information included self-reported smoking status, whether women received any blood transfusions over gestation and if prenatal vitamins were prescribed during antenatal visits. Smoking status was reported as “never smoked”, “passive smoking”, 50 “quit smoking” or “currently smoking” based on information present in the medical records. In addition, data were also abstracted on gestational age at delivery, number of prenatal care visits, parity, multiple birth delivery, use of assisted reproduction technologies, delivery type (vaginal, C-section, vaginal birth after a cesarean section (VBAC), or other). Adequacy of prenatal care was determined using the Adequacy of Prenatal Care Utilization Index which was evaluated based on the timing of entry into prenatal care, total number of prenatal visits, and gestational age at delivery using Kotelchuck’s method (18). Race was self-reported and classified into four categories as Asian (Asian, Asian Indian, Bangladeshi, Bhutanese, Burmese, Cambodian, Chinese, Filipino, Indonesian, Japanese, Korean, Laotian, Nepalese, Pakistani, Sri Lankan, Thai, Vietnamese), Black (Black or African American), White (White or Caucasian), or Other (American Indian or Alaskan Native, Native Hawaiian or Other Pacific Islander, Other, Other Pacific Islander, Samoan, Solomon Islander, Tahitian, Unknown or Patient Refused). Ethnicity was self-reported and classified as Hispanic, Non-Hispanic or Unknown. All available Hb and SF data obtained across gestation were extracted from medical records. Hemoglobin analyses at the Strong Memorial Hospital and Highland Hospital were undertaken using the Cell-Dyn 4000 system (Abbott, Santa Clara, CA) in a CLIA certified laboratory. Serum ferritin concentrations were measured by Roche Cobas e601/e602 (Roche Diagnostics Rotkreuz, Switzerland). For women that had multiple Hb measurements obtained within the same trimester of pregnancy, the mean Hb concentration of the multiple values within each trimester was used for analysis purposes and the mean gestational age of the replicate measures was used as the time of measure. For women who had two Hb measurements (n = 154) 51 on the delivery date, the higher Hb value was used in data analyses. Anemia was defined using the CDC criteria as a Hb concentration < 11.0 g/dL in T1 or T3, or < 10.5 g/dL during T2. Data were also evaluated using race adjusted cutoffs to define anemia in Black women, whereby cut-offs were lowered by 0.8 g/dL as indicated by the CDC (19). Analyses were undertaken with and without this race adjustment factor. We also examined outcomes associated with elevated Hb concentration defined as either a Hb concentration greater than 13.0 g/dL (as Hb concentrations above this value have been associated with adverse health outcomes (2), or defined as a Hb concentration greater than 15.0 g/dL, as this was the 95th percentile of Hb concentrations based on data from the CDC reference population (19). Statistical analyses Descriptive statistics were used to summarize characteristics of pregnant women, as the mean ± standard deviation (SD), or median (interquartile range (IQR)) for continuous variables or counts and percentages for categorical variables. The distribution of Hb concentrations in each trimester was examined. Histograms were used to visualize the Hb distribution across each trimester and smoothed lines were fit to characterize distributions. Mean and standard deviation (SD) of Hb and the 3rd, 5th, 95th and 97th percentiles of Hb concentrations in each trimester were calculated. Possible differences in Hb concentrations between trimesters or sub-groups of the population were evaluated using mixed-effects linear regression models setting mother as the random effect to account for women that contributed more than one pregnancy over the 10-year study period. Least-square means were calculated in the post-hoc analyses when evaluating pairwise comparisons between trimesters. The overall p-values of the mixed-effects linear 52 regression models were calculated by F-tests. Possible differences in the prevalence of anemia and elevated Hb concentrations between trimesters were examined by mixed-effects logistic regression models. To identify risk factors associated with anemia and elevated Hb, mixed effects multiple logistic regression models were utilized controlling for trimester of pregnancy. The overall p-values of the mixed effects logistic regression models were obtained using Wald chi-square tests. Risk of anemia was evaluated using odds ratios (OR) with a 95% confidence interval. Statistical significance was defined as P < 0.05. All statistical analyses were conducted using R (version 4.0.3). Results Characteristics of the study population Over the 10-year period from 2011-2020 a total of 54,453 pregnancies occurred. Of these, pregnancies that did not end in a live birth (n = 823), pregnancies that occurred to women diagnosed with diabetes (both gestational diabetes or type 1 or type 2 diabetes) (n = 3,829), autoimmune diseases (n = 121), HIV infection (n = 64), gastrointestinal disorders (n = 624), hemoglobinopathies (n = 385), hypertension including eclampsia and preeclampsia (n = 7,838), and thrombocytopenia (n = 130) were excluded resulting in a database that included 42,117 pregnancies that occurred to 31,158 women. The final data set was further cleaned to eliminate datapoints that were deemed to be biologically implausible and were attributed to data entry errors. These included Hb concentrations greater than 20.0 g/dL (n = 5) or less than 5.0 g/dL (n = 53 7), body weight < 35.0 kg (n = 16) or > 300.0 kg (n = 6), or height >2.0 m (n = 8) or < 1.2 m (n = 16). After further excluding healthy pregnancies without any Hb data available in the medical chart (n = 891), the final analytic sample included 41,226 births that occurred in 30,603 women (Figure 2.1). Of all 54,453 pregnancies occurred between 2011-2020, 5% (n = 2,752) had no available Hb data, and of the 42,117 uncomplicated pregnancies, 2% (n = 888) had no Hb data in the medical record. Figure 2.1 Flow chart of the study population identified from the retrospective medical chart review. 54 For pregnancies remaining in the final analytic sample, an average of 2.5 Hb measures were obtained across gestation. Of the Hb measures obtained, 28.6% were obtained in T1, 14.4% were obtained during T2 and 57.0% were obtained during T3. In our cohort, 78% of pregnancies had Hb measured in more than 2 trimesters: 4821 (12%) pregnancies had Hb measured in all three trimesters and 27,300 (66%) pregnancies had Hb measures in two trimesters. Of pregnancies that had longitudinal Hb measures (n = 32,121) in different trimesters of pregnancy, 0.5% (n = 168) had Hb measures in T1 and T2, 63.3% (n = 20,331) had Hb measures in T1 and T3, 21.2% (n = 6,801) had Hb measures in T2 and T3, and 15.0% (n = 4,821) had Hb measures in all three trimesters. Only 2.3% of the final pregnancy sample (n = 948) had SF data in their medical record. When obtained, the majority of SF analyses (41.2%) were obtained during T3, followed by T1 (30.2%) and T2 (28.5%). Among pregnancies with available SF data, 20% (n = 192) had more than one SF measurement obtained across pregnancy. Characteristics of women in this study population are presented in Table 2.1. Only a small percentage of pregnant women receiving care were adolescents (≤ 19 y of age). The majority of women self-identified their race/ethnicity as White and non-Hispanic. The mean pre- pregnancy BMI fell slightly over the normal weight category, but the majority of women (74.5%) had missing pre-pregnancy BMI data in their medical chart (Table 2.1). Among all pregnancies with pre-pregnancy BMI data available, 44.4% of women were of normal BMI, 3.3% were underweight, 25.2% were overweight, 21.3% were obese and 5.9% were severely obese (≥ 40.0 kg/m2) (data not shown). Almost half of the pregnancies with a known number of prenatal care visits (46.7%) were categorized as having Inadequate prenatal care based on the 55 Adequacy of Prenatal Care Utilization Index (Table 2.1). After preliminary analyses, the “currently smoking” (n = 3,097) and “passive smoking” (n = 180) groups had Hb concentrations that did not significantly differ and thus these were grouped together in the following analyses to increase the statistical power. Table 2.1 Characteristics of pregnant women Variable Groups Mean (SD) or N (%) Median (IQR) Age (y) 29.8 (5.5) * 41,226 (100) Adolescents < 20 y 18 (2) 1,405 (3.4) Adults ≥ 20 y 30.2 (5.2) * 39,821 (96.6) Race - Asian - 1,594 (3.9) Black - 8,325 (20.2) White - 27,417 (66.5) Other - 3,890 (9.4) Ethnicity - Hispanic - 3,556 (8.63) Non-Hispanic - 36,061 (87.5) Unknown - 1,609 (3.90) Pre-pregnancy BMI (kg/m2) 25.4 (8.7) 10,507 (25.5) Underweight (< 18.5) 17.8 (1.0) 350 (0.9) Normal weight (18.5 - < 25) 22.0 (1.7) * 4,665 (11.3) Overweight (25.0 - < 30) 27.3 (1.4) * 2,643 (6.4) Obese (30.0 - < 40) 34.1 (2.8) * 2,233 (5.4) Severe Obesity (>= 40.0) 43.8 (5.4) 616 (1.8) Missing BMI data - 30,719 (74.5) Gestational weight gain (kg) 8.7 (5.7) * - Underweight (< 18.5) 10.0 (5.3) - Normal weight (18.5 - < 25) 9.8 (5.4) - 8.9 (5.3) * - Overweight (25.0 - < 30) Obese (30.0 - < 40) 6.3 (6.2) * - Severe Obesity (>= 40.0) 3.6 (8.2) - Parity 1 (2) 41,174 (99.9) Nulliparous (Parity = 0) 0 (0) 16,330 (39.6) Primiparous (Parity = 1) 1 (0) 14,056 (34.1) Multiparous (Parity > 1) 2 (1) 10,788 (26.2) Unknown - 52 (0.1) 56 Table 2.1 (Continued) Use of assisted reproduction technology Yes - 271 (0.7) No - 40,955 (99.3) Blood transfusion during pregnancy - Yes - 588 (1.4) No 40,638 (98.6) Gestational weeks at delivery 39 (2) 41,226 (100) Preterm (< 37 weeks) 34 (4) 3,533 (8.6) Not preterm (≥ 37 weeks) 39 (1) 37,693 (91.4) Multiple pregnancy - No - 40,278 (97.7) Yes - 920 (2.2) Unknown - 28 (0.1) Delivery type - Vaginal - 29,593 (71.8) C-section - 10,801 (26.2) VBAC - 442 (1.1) Other - 38 (0.1) Unknown - 354 (0.8) Adequacy of Prenatal Care Utilization Index Inadequate - 6,640 (16.1) Intermediate - 1,000 (2.4) Adequate - 3,365 (8.2) Adequate Plus - 3,202 (7.8) Unknown 27,019 (65.5) Vitamins prescribed during pregnancy - Yes - 14,346 (34.8) No - 26,880 (65.2) Smoking status - Never - 27,464 (66.6) Passive or Yes - 3,277 (8.0) Quit - 8,033 (19.5) Not Asked or Unknown - 2,452 (5.9) * Data are presented as Median (IQR). VBAC, vaginal birth after a cesarean section. Hemoglobin distributions among pregnant women Table 2.2 summarizes Hb data, and the prevalence of anemia or elevated Hb in each trimester. Using all available Hb data, mean Hb averaged 12.0 g/dL; the prevalence of anemia 57 and Hb > 13 g/dL or Hb >15g/dL was 16.7%, 20.2% and 0.3% respectively. Because there were too few pregnant women with Hb concentrations greater than 15.0 g/dL, statistical analyses using this definition of elevated Hb were not used in further group analyses. The mean non-race adjusted Hb concentrations were significantly higher in T1 (P < 0.0001), compared to both T2 and T3 but Hb concentrations did not significantly differ between T2 and T3 (P = 0.4). The prevalence of anemia increased significantly across pregnancy, being highest in T3 (P < 0.0001) (Table 2.2). When anemia classifications were adjusted for race, the prevalence of anemia decreased by approximately 4 percentage points (i.e. from 16.7% to 12.8%). The 5th percentiles of the observed Hb concentration in the group as a whole at T1, T2, and T3 was 11.0 g/dL, 9.8 g/dL and 9.6 g/dL respectively. The recommended CDC anemia cutoffs selected as the 5th percentile of their reference population represented the 5th, 16th and 27th percentile of our data distribution in each respective trimester (Figure 2.2). 58 Figure 2.2 Hemoglobin distributions across gestation. Blue lines reflect the 5th percentile of the CDC reference data, red lines reflect the 5th percentile of our study population. The 95th percentile (15.0 g/dL) of the CDC reference Hb distribution used to define an elevated Hb concentration was above the 99th percentile of the observed Hb concentration in our cohort. The highest observed Hb concentration was found at 4 weeks of gestation (13.1 g/dL) and the nadir of the Hb distribution was found at 30 weeks of gestation (11.0 g/dL). Values rose from 30 weeks of gestation until 40 weeks of gestation (Figure 2.3). 59 Figure 2.3 Hemoglobin concentrations across gestation. Dots reflect the mean hemoglobin concentration observed at each gestational week presented for each racial group. Error bars reflect the 95% confidence intervals of the mean. Table 2.2 Hemoglobin concentrations and anemia/elevated hemoglobin prevalence across gestation Total T1 T2 T3 P-value1 Hb concentrations (g/dL) < 0.0001 Mean 12.0 12.6 11.6 11.8 (SD) (1.2) (1.0) (1.1) (1.3) Min 5.4 5.9 6.1 5.4 Max 18.6 18.6 16.9 16.6 3rd %ile 9.5 10.7 9.4 9.2 5th %ile 9.9 11.0 9.8 9.6 Median 12.1 12.7 11.6 11.8 95th %ile 13.9 14.1 13.3 13.8 97th %ile 14.1 14.3 13.6 14.1 Anemia prevalence % 16.7 4.5 14.0 25.1 < 0.0001 Anemia prevalence (race adjusted) % 12.8 2.9 8.6 20.3 < 0.0001 Elevated Hb (Hb > 13 g/dL) % 20.2 32.9 8.2 15.7 < 0.0001 Elevated Hb (Hb > 15 g/dL) % 0.3 0.4 0.1 0.3 - 60 1 Differences between hemoglobin concentrations among trimesters were analyzed by F-tests. Differences between anemia prevalence or elevated hemoglobin between trimesters were analyzed by Wald chi- squared tests. T1= first trimester; T2 = second trimester; T3 = third trimester. Hemoglobin concentration comparisons between subgroups of study population Hemoglobin concentration comparisons between sub-groups are presented in Table 2.3. Adult pregnant women had significantly higher Hb concentrations (+0.6 g/dL, P < 0.0001) than observed among pregnant adolescents. Mean Hb concentrations were significantly lower among Black women in each trimester on average by -0.8 (T1), -0.7 (T2), and -1.0 g/dL (T3) when compared to White women (P < 0.0001). White women had the highest Hb concentrations in T1 and T2 (12.8 ± 0.9, 11.9 ± 1.0 g/dL), while Asian women had the highest Hb concentrations in T3 (12.2 ± 1.2 g/dL, n = 1,577, P < 0.0001) (Figure 2.3). Black women had lowest Hb concentrations in all trimesters compared to other racial groups (12.0 ± 1.1, 11.2 ± 1.1, and 11.0 ± 1.3g/dL, P < 0.0001). In addition, non-Hispanic women (n = 3,556) had significantly higher Hb concentrations when compared to Hispanic women (n = 36,061) across gestation (P < 0.0001). Pregnant women with a normal or overweight ppBMI (n = 7,308) had significantly higher Hb concentrations (P < 0.0001) compared to means observed in underweight or obese women (n = 3,199). Pregnant women who were classified as receiving Inadequate prenatal care had significantly higher Hb concentrations (P < 0.0001) than other groups (Table 2.3). Interestingly, there were more White, Asian women (P < 0.0001), more non-Hispanic women (P < 0.0001), more nulliparous women (P < 0.0001) and more adolescents (P = 0.0013) were categorized as receiving Inadequate prenatal care. Pregnant women who stated that they never 61 smoked had significantly (P < 0.0001) higher Hb concentrations than women who smoked during pregnancy (Table 2.3). Table 2.3 Mean hemoglobin concentrations as a function of key study variables Hb (g/dL) Overall P-value1 Age (y) < 0.0001 Adolescent (< 20 y) 11.4 (1.2) a Adult (≥ 20 y) 12.0 (1.2) b Race < 0.0001 Asian 12.2 (1.1) a Black 11.3 (1.2) b White 12.3 (1.1) c Other 11.8 (1.2) d Ethnicity < 0.0001 Hispanic 11.8 (1.2) a Non-Hispanic 12.0 (1.2) b Pre-pregnancy BMI (kg/m2) < 0.0001 Underweight (< 18.5) 11.6(1.4) ab Normal weight (18.5 - < 25) 11.9 (1.3) c Overweight (25.0 - < 30) 11.9 (1.3) c Obese (30.0 - < 40) 11.8 (1.2) a Severe Obese (>= 40.0) 11.6 (1.4) b Parity < 0.0001 Nulliparous (Parity = 0) 12.2 (1.2) a Primiparous (Parity = 1) 12.1 (1.2) b Multiparous (Parity > 1) 11.7 (1.3) c Blood transfusion during pregnancy < 0.0001 Yes 11.0 (1.8) a No 12.0 (1.2) b Gestational weeks at delivery < 0.0001 Preterm (<37 weeks) 11.7 (1.3) a Not preterm (≥ 37weeks) 12.1 (1.2) b Multiple pregnancy < 0.0001 Singleton 12.0 (1.2) a Multiples 11.6 (1.4) b Delivery type < 0.0001 Vaginal 12.1 (1.2) a C-section 12.0 (1.3) b VBAC 11.9 (1.3) ab Other 11.9 (1.2) ab 62 Table 2.3 (Continued) Adequacy of Prenatal Care Utilization Index < 0.0001 Inadequate 12.0 (1.3) a Intermediate 11.7 (1.3) b Adequate 11.9 (1.2) b Adequate Plus 11.8 (1.3) b Vitamins prescribed during pregnancy < 0.0001 Yes 11.9 (1.3) a No 12.1 (1.3) b Smoking status < 0.0001 Never 12.1 (1.2) a Quit 11.9 (1.2) b Yes or Passive 11.7 (1.3) c Data are presented as mean  SD. Different superscripts indicate a significant difference in hemoglobin concentrations between sub-groups using mixed linear regression models. VBAC, vaginal birth after a cesarean section. 1 Possible difference in hemoglobin concentrations between trimesters were evaluated using F-tests. Anemia and high hemoglobin prevalence comparisons between subgroups of study population In the group as a whole, nearly 1 in 5 pregnant women were anemic at one point over the course of gestation. Not surprisingly, factors that were associated with the prevalence of anemia (Table 2.4) were similar to the factors found to be associated with maternal Hb concentrations. The prevalence of anemia was highest among Black women in each trimester of pregnancy (13.6%, 24.2% and 47.6%), while White women had the lowest prevalence of anemia in the T1 (2.0%) and T2 (7.9%), Asian women had the lowest prevalence of anemia in T3 (15.3%). The prevalence of anemia increased by 8-fold from T1 (2.0%) to T3 (18.3%) in White women, by 63 2.5-fold (from 13.6% to 47.6%) in Black women, while only by 1.7-fold in Asian women (from 5.7% to 15.3%) (Figure 2.4). Figure 2.4 Self-reported racial groups and anemia prevalence. (*) indicates significant difference in anemia prevalence among racial groups in each trimester (P < 0.0001). (a, b, c) indicate significant difference in anemia prevalence among trimesters (P < 0.0001). After race-adjusting the anemia cut-off for Black women, the prevalence of anemia remained significantly higher among Black women compared to Asian women and White women (P < 0.001) (Table 2.4). Women who were classified as underweight or severely obese based on ppBMI were 2 times more likely to be anemic compared to women with normal or overweight ppBMI (P < 0.001). Prevalence of obesity or severe obesity was 2 times higher among Black women compared to White women and prevalence of being underweight was 2.4 times higher among Asian women using ppBMI. Both ppBMI category and race were independently significantly associated with anemia. Multiparous and primiparous women were 4.6 and 2.3 64 times more likely to have anemia compared to nulliparous pregnant women (P < 0.0001). Women who were carrying a multiple pregnancy had almost 6-fold increased risk of gestational anemia compared to women carrying a singleton pregnancy (P < 0.0001). Unexpectedly, pregnant women who either smoked during pregnancy or were exposed to secondhand smoke were 2 times more likely to be anemic compared to women who had never smoked cigarettes (P < 0.0001). The comparisons of prevalence of elevated Hb between sub-groups are shown in the Table 2.5. Table 2.4 Prevalence of anemia as a function of key study variables Anemia Pairwise Overall Anemia (race Pairwise Overall (%) P-value P-value adjusted) (%) P-value P-value Age (y) 0.80 0.17 Adolescent (< 20 y) 30.2 - 19.5 - Adult (≥ 20 y) (Ref) 16.2 0.80 12.6 0.09 Race <0.0001 <0.0001 Asian 11.3 0.99 11.3 1.0 Black 32.9 <0.0001 14.5 0.0009 White (Ref) 11.3 - 11.3 - Other 19.7 <0.0001 19.7 <0.0001 Ethnicity <0.0001 <0.0001 Hispanic 21.6 <0.0001 20.3 - Non-Hispanic (Ref) 16.3 - 12.1 <0.0001 Pre-pregnancy BMI (kg/m2) < 0.0001 0.23 Underweight (< 18.5) 26.9 0.0004 20.2 0.18 Normal weight (18.5 - < 25) (Ref) 18.7 - 14.1 - Overweight (25.0 - < 30) 20.3 0.14 14.1 1.0 Obese (30.0 - < 40) 20.8 0.008 13.1 1.0 Severely Obese (>= 40.0) 26.4 <0.0001 15.3 0.92 Parity <0.0001 <0.0001 Nulliparous (Parity = 0) (Ref) 13.0 - 9.8 - Primiparous (Parity = 1 15.9 <0.0001 12.7 <0.0001 Multiparous (Parity > 1) 23.0 <0.0001 17.2 <0.0001 Multiple pregnancy < 0.0001 <0.0001 Singleton (Ref) 16.4 - 12.5 - Multiples 26.6 <0.0001 23.8 <0.0001 65 Table 2.4 (Continued) Adequacy of Prenatal Care Utilization Index <0.0001 0.10 Inadequate 19.5 <0.0001 14.3 0.17 Intermediate 25.2 0.29 17.2 0.98 Adequate (Ref) 20.4 - 13.8 - Adequate Plus 20.2 0.96 13.9 0.38 Smoking status <0.0001 <0.0001 Never (Ref) 15.7 - 12.1 - Quit 18.0 0.0006 13.7 <0.0001 Yes or Passive 23.4 <0.0001 16.9 0.02 Differences between anemia prevalence among sub-groups were analyzed by mixed logistic model from Wald chi-squared tests. (Ref) indicates the reference group for each group comparison. Table 2.5 Prevalence of elevated hemoglobin concentrations (> 13.0 g/dL) as a function of key study variables Elevated Hb Pairwise Overall (Hb > 13) (%) P-value P-value Age (y) < 0.0001 Adolescent (< 20 y) 8.9 - Adult (≥ 20 y) (Ref) 20.6 < 0.0001 Race <0.0001 Asian 21.1 0.0005 Black 6.8 <0.0001 White (Ref) 25.1 - Other 15.3 <0.0001 Ethnicity <0.0001 Hispanic 13.4 <0.0001 Non-Hispanic (Ref) 20.8 - Pre-pregnancy BMI (kg/m2) < 0.0001 Underweight (< 18.5) 13.8 0.49 Normal weight (18.5 - <25) (Ref) 17.9 - Overweight (25.0 - < 30) 16.9 0.80 Obese (30.0 - < 40) 13.9 <0.0001 Severe Obese (>= 40.0) 9.8 <0.0001 Parity <0.0001 Nulliparous (Parity = 0) (Ref) 23.1 - Primiparous (Parity = 1 21.0 <0.0001 Multiparous (Parity > 1) 14.8 <0.0001 Multiple pregnancy < 0.0001 Singleton (ref) 20.3 - Multiples 13.5 <0.0001 66 Table 2.4 (Continued) Adequacy of Prenatal Care Utilization Index <0.0001 Inadequate 18.3 0.0002 Intermediate 13.6 1.00 Adequate (ref) 16.2 - Adequate Plus 16.4 0.72 Smoking status <0.0001 Never (ref) 21.0 - Quit 18.2 <0.0001 Yes or Passive 15.0 <0.0001 Differences between elevated hemoglobin prevalence among sub-groups were analyzed using mixed logistic models from Wald chi-squared tests. (Ref) indicates the reference group for each group comparison. Determinants of anemia during pregnancy Determinants of anemia are presented in Table 2.6. Because 75% of pregnancies had missing ppBMI data, models were constructed including and excluding ppBMI as an independent variable. When maternal ppBMI was included in the model as an independent variable, variables found to be associated with a significantly higher odds ratio for anemia included Hb measurement during T2 or T3 of pregnancy; adolescent mother; Black or Other racial group, Hispanic ethnicity; being underweight; higher parity, carrying multiple fetuses, and smoking/passive smoking during pregnancy. In models that used race adjusted anemia cutoffs, maternal age, being of Black race or Hispanic ethnicity, and self-reported history of smoking during pregnancy were no longer statistically associated with anemia. When maternal ppBMI was excluded in the models, findings were similar to the results obtained using ppBMI, but pregnant adolescents and those who smoked during pregnancy had a significantly higher odds ratio for anemia using either the race-adjusted or non-race adjusted anemia cutoffs. The smoking 67 status association with Hb contradicts some findings on the impact of smoking on Hb concentrations reported in other studies carried out in non-pregnant adults. Cigarette smoking has been found to right shift the Hb distribution curve and Hb concentrations are often elevated in smokers compared with non-smokers as a response to the hypoxia caused by smoking (26). However, several previous studies in pregnant women have reported that smoking was associated with decreased Hb levels and increased risk of IDA (27–30). Although anemia risk factors including multiparity, underweight and being of Black race were associated with higher prevalence of smoking during pregnancy in our cohort, the odds ratio for anemia was still significantly higher in smokers after controlling for these variables. It is possible that other risk factors for anemia during pregnancy have a more pronounced effect on Hb concentrations, but our study was not designed to answer this question. 68 Table 2.6. Determinants of anemia ppBMI included as an independent variable (n = 10,279) ppBMI not included as an independent variable (n = 37,416) Anemia (R2 = 0.67) Adjusted anemia Anemia (R2 = 0.77) Adjusted Anemia (R2 = 0.98) (R2 = 0.97) Variables OR (95% CI) P-value OR (95% CI) P-value OR (95% CI) P-value OR (95% CI) P-value Trimesters First Ref - Ref - Ref - Ref - Second 2.8 (2.3, 3.3) < 0.0001 7.1 (5.2, 9.7) < 0.0001 3.5 (3.1, 3.98) < 0.0001 6.4 (5.4, 7.7) < 0.0001 Third 20.6 (17.3 24.7) < 0.0001 1,603.8 < 0.0001 26.1 (22.5, 30.4) < 0.0001 386.0 < 0.0001 (1,128.6, 2,279.0) (324.4, 459.4) Maternal age at delivery (y) Adolescent (< 20 y) 1.9 (1.4,2.4) < 0.0001 1.6 (0.8, 2.8) 0.13 2.0 (1.6, 2.4) < 0.0001 1.7 (1.3, 2.4) 0.0008 Adult (≥ 20 y) Ref - Ref - Ref - Ref - Race Asian 1.1 (0.8, 1.7) 0.52 1.1 (0.5, 2.5) 0.88 1.2 (0.9, 1.5) 0.15 1.2 (0.8, 1.8) 0.35 Black 7.4 (6.2 8.8) < 0.0001 1.0 (0.7, 1.4) 0.94 13.2 (11.1, 15.7) < 0.0001 1.2 (1.0, 1.5) 0.05 White Ref Ref Ref Ref Other 2.2 (1.7, 2.9) < 0.0001 2.0 (1.16, 3.7) 0.02 2.4 (2.0, 2.9) < 0.0001 2.0 (1.4, 2.8) < 0.0001 Ethnicity Hispanic 1.3 (1.0, 1.7) 0.02 1.4 (0.8, 2.4) 0.21 1.7 (1.4, 2.1) < 0.0001 1.7 (1.2, 2.3) 0.0009 Non-Hispanic Ref Ref Ref Ref Pre-pregnancy BMI (kg/m2) Underweight (< 18.5) 1.7 (1.2, 2.3) 0.003 2.2 (1.1, 4.4) 0.03 - - - - Normal weight (18.5 - < 25) Ref - Ref - - - - - Overweight (25.0 - < 30) 1.0 (0.8, 1.1) 0.83 0.9 (0.6, 1.2) 0.35 - - - - Obese (30.0 - < 40) 0.8 (0.7, 0.9) 0.006 0.7 (0.5, 1.1) 0.13 - - - - Severe Obese (>= 40.0) 1.2 (0.9, 1.5) 0.30 0.9 (0.5, 1.7) 0.79 - - - - Parity Nulliparous (Parity = 0) Ref Ref Ref Ref Primiparous (Parity = 1 1.9 (1.6, 2.2) < 0.0001 2.9 (2.2, 4.0) < 0.0001 1.8 (1.7, 2.0) < 0.0001 2.6 (2.3, 3.0) < 0.0001 Multiparous (Parity > 1) 2.8 (2.4, 3.3) < 0.0001 4.7 (3.4, 6.6) < 0.0001 3.0 (2.7, 3.4) < 0.0001 4.6 (3.9, 5.4) < 0.0001 69 Table 2.6 (Continued) Multiple pregnancy Singleton Ref Ref Ref Ref Multiples 3.8 (2.6, 5.6) < 0.0001 16.3 (6.3, 42.1) < 0.0001 3.9 (3.1, 5.0) < 0.0001 7.2 (5.0, 10.5) < 0.0001 Smoking status Never Ref Ref Ref Ref Quit 1.0 (0.9, 1.2) 0.64 1.0 (0.7, 1.4) 0.90 1.2 (1.1, 1.3) 0.001 1.2 (1.0, 1.4) 0.11 Yes or Passive 1.3 (1.0, 1.6) 0.03 1.5 (1.0, 2.4) 0.07 1.6 (1.3, 1.8) < 0.0001 1.5 (1.2, 2.0) 0.0006 ORs indicate adjusted odds ratio of anemia analyzed by the mixed effect multiple logistic regression models. (Ref) indicates the reference group for each group comparison. ppBMI, pre-pregnancy body mass index. 70 Discussion This retrospective medical chart review of 41,226 pregnancies that occurred in a group of 30,603 healthy US pregnant women provided cross-sectional and longitudinal data on Hb concentrations across gestation. Anemia was evident in 16.7 % of these otherwise healthy pregnant women receiving standard prenatal care. The prevalence of anemia increased as pregnancy progressed, such that by late pregnancy fully 25% of women evaluated for anemia during the third trimester of pregnancy were anemic. Notably, the current CDC anemia cutoffs based on the 5th percentile of the reference population in each trimester of pregnancy actually represented the 5th, 16th and the 27th percentiles of our population in T1, T2 and T3 respectively. Risk factors for anemia included Black race or Hispanic ethnicity, being underweight at entry into pregnancy, being multiparous or carrying a multiple birth pregnancy. The prevalence of anemia across gestation was lowest among White and Asian women, and risk of elevated Hb concentrations was highest among White and Asian women. The overall prevalence of anemia in this large cohort of women was 2 – 6.5 times higher than previously reported using national NHANES survey data (5–7). Evaluating data by trimester, the prevalence of anemia in each trimester remained more than 2 times higher than reported in the 1999-2006 NHANES survey (7). Although the 1999-2010 NHANES data only included singleton pregnancies, their pregnant population included a higher percentage of adolescents (6.3% vs. 3.4%) and a lower percentage of White women (53.5% vs. 66.5%) compared to our study population. In spite of having a greater percentage of women with these 71 two risk factors that we have identified as being associated with increased risk of anemia, the anemia prevalence they reported was markedly higher than that observed in our far larger cohort of women. The reasons for these discrepancies are unknown but may be impacted by the relatively small size of the NHANES study population (n = 1,283). Notably, the current CDC anemia cutoffs that were identified as the 5th percentile of the reference population in each trimester of pregnancy reflected the 5th, 16th and the 27th percentiles of this much larger obstetric population in T1, T2 and T3 respectively. Recent Hb data obtained using pooled data from 257 population-representative data sources from 107 countries worldwide reported an average Hb of 11.4 g/dL in the group as a whole and from the figure presented in their paper it appears that the 5th percentile of their Hb distribution was also markedly lower than reported in the CDC data (Figure 2.5) (15). Figure 2.5 Global distributions of hemoglobin concentration for pregnant and non-pregnant women in 2011 presented by Steven et al. 72 The results of a recent international study presenting the Hb distributions in 3,502 healthy women pregnant carrying singletons also reported findings comparable to ours in that their reported 5th centile value at T2 and T3 (from 9.7 to 10.4 g/dL) were also lower than those currently utilized by the CDC (16). Hemoglobin concentrations higher than 9.5 g/dL-11.0 g/dL have found to be associated with higher risks of preeclampsia, prematurity, and fetal growth restriction (21), thus the current CDC reference anemia cutoffs may not be representative of current Hb distributions and if so this would overestimate the prevalence of anemia. Interestingly, the 5th percentile of our cohort was the same as the CDC reference population in T1, but the Hb distribution shifted to the left in T2 and shifted to a larger extent in T3 in our population. One possible explanation might relate to the much larger dose of supplemental iron ingested by the CDC reference cohort. The European populations used to generate the CDC reference data ingested between 65 to 200 mg of supplemental iron per day which 2-7 times higher than the current US Recommended Dietary (27 mg/d) (14). In these European studies, the authors found that pregnant women receiving these high doses of oral iron supplementation during pregnancy maintained their iron stores and maintained normocytic erythropoiesis through T3, while compromised erythropoiesis was evident in non-iron supplemented women (10,11,13). We speculate that the current lower doses of iron supplementation may lead to the left shifted Hb distribution observed in our cohort. The CDC reference population was obtained in a cohort of women who were not representative of the current US pregnant population in regard to race/ethnic composition, ppBMI and supplemental iron intake (14). Larger more diverse cohorts of women that better resemble the racial and ethnic 73 composition and ppBMI distribution of the current US obstetric population, and that are ingesting supplemental iron intakes consistent with recent recommendations are needed to characterize anemia cutoffs during pregnancy and to help identify the current prevalence of anemia among US women. Late gestation, Black race, multiple gestation pregnancy and multiparity were the strongest predictors of anemia during pregnancy. The anemia prevalence increased 3-fold across gestation, which is comparable to that reported in the NHANES 1999-2006 data (7). Racial disparities in Hb, with greater risk of anemia among women who self-identify as Black, have been identified in many previous epidemiological and population studies (5–7,22) and some CDC guidelines have suggested lowering the anemia cutoffs for Black women by 0.8 g/dL (19). The relative difference in Hb concentrations in our cohort of women are comparable to the CDC data as mean Hb concentrations were 0.8 g/dL lower in Black pregnant women compared to White women. Whether this difference in Hb concentration is a consequence of physiological changes or to differences in iron status requires further investigation and use of these race- specific cut-offs should be interpreted with caution in clinical settings to avoid the risk of under- diagnosing anemic women. We also noted that Asian pregnant women had the lowest prevalence of anemia across gestation, and the highest risk of elevated Hb concentrations in late gestation. Few data have evaluated Hb changes across gestation in Asian populations but the largest epidemiological study to date that evaluated iron stores as a function of ethnicity, the Hemochromatosis and Iron Overload Screening (HEIRS) Study, found Asians had the highest risk of elevated iron stores (23). Among women enrolled in the HEIRS study that were studied 74 while pregnant or breastfeeding, Asian ethnicity was also found to be associated with a decreased risk of ID (24). In agreement with these epidemiologic data, iron absorption has been reported to be significantly higher among non-pregnant East Asian women when compared to White women, even after controlling for iron stores as evaluated using SF (25). More data investigating the longitudinal changes in Hb across pregnancy and the impact of ethnicity on iron homeostasis are needed. This study has some limitations. No data on birth outcomes were available and we are unable to address the impact of gestational anemia or elevated Hb concentrations during pregnancy on adverse maternal or neonatal birth outcomes. Data on maternal educational level and socioeconomic status were not available to evaluate the potential impact of these factors on risk of anemia or to compare the characteristics of this cohort to other published data on determinants of anemia. These pregnant women resided in Rochester, NY and the surrounding community (elevation of 214 meters above sea level) and these data would likely not be relevant for those living at high altitudes. Our study provides the largest normative longitudinal datasets on Hb across pregnancy in otherwise healthy US pregnant women receiving routine prenatal care. These results add to the current knowledge on normative hemoglobin concentrations in US pregnant women and help identify determinants of anemia across pregnancy. We can conclude from this medical chart review that anemia was prevalent in otherwise healthy US pregnant women receiving standard prenatal care even with the current universal iron supplementation recommendations. The prevalence of anemia increased as pregnancy progressed. Risk factors for anemia included Black 75 race, being underweight at entry into pregnancy, being multiparous or carrying a multiple birth pregnancy. The hemoglobin distribution in the population as a whole was left-shifted in T2 and T3 compared to the CDC reference population but was similar to other recent global studies of hemoglobin in large pregnancy cohorts (15,16). Additional studies are needed to evaluate the impact of maternal anemia on iron status in the newborn at birth and its association with maternal birth outcomes to evaluate the current hemoglobin cut-offs used to define anemia. Targeted identification of women at greater risk may help inform interventions to improve maternal iron status and the in-utero environment in support of healthy birth outcomes. 76 REFERENCES 1. World Health Organization. Worldwide prevalence of anaemia 1993–2005. WHO Global Database on Anaemia. WHO Rep. 2005;51. 2. Young MF, Oaks BM, Tandon S, et al. Maternal hemoglobin concentrations across pregnancy and maternal and child health: a systematic review and meta-analysis. Ann N Y Acad Sci. 2019;1450(1):47. 3. Rahman MM, Abe SK, Rahman MS, et al. Maternal anemia and risk of adverse birth and health outcomes in low- and middle-income countries: Systematic review and meta- analysis. Am J Clin Nutr. 2016;103(2):495–504. 4. Daru J, Zamora J, Fernández-Félix BM, et al. Risk of maternal mortality in women with severe anaemia during pregnancy and post partum: a multilevel analysis. Lancet Glob Heal. 2018;6(5):e548–e554. 5. Gupta PM, Hamner HC, Suchdev PS, Flores-Ayala R, Mei Z. Iron status of toddlers, nonpregnant females, and pregnant females in the United States. Am J Clin Nutr. 2017;106(Suppl 6):1640S-1646S. 6. Le CHH. The prevalence of anemia and moderate-severe anemia in the US population (NHANES 2003-2012). PLoS One. 2016;11(11):e0166635. 7. Mei Z, Cogswell ME, Looker AC, et al. Assessment of iron status in US pregnant women from the national health and nutrition examination survey (NHANES), 1999-2006. Am J Clin Nutr. 2011;93(6):1312–1320. 8. Siu AL, Albert L. Siu. Screening for Iron Deficiency Anemia and Iron Supplementation in 77 Pregnant Women to Improve Maternal Health and Birth Outcomes: U.S. Preventive Services Task Force Recommendation Statement. American College of Physicians; 2015. 9. Puolakka J, Jäne O, Pakarinen A, Järvinen PA, Vihko R. Serum Ferritin as a Measure of Iron Stores During and After Normal Pregnancy with and Without Iron Supplements. Acta Obstet Gynecol Scand. 1980;59(95 S):43–51. 10. TAYLOR DJ, MALLEN C, McDOUGALL N, LIND T. Effect of iron supplementation on serum ferritin levels during and after pregnancy. BJOG An Int J Obstet Gynaecol. 1982;89(12):1011–1017. 11. Sjöstedt JE, Manner P, Nummi S, Ekenved G. ORAL IRON PROPHYLAXIS DURING PREGNANCY-A COMPARATIVE STUDY ON DIFFERENT DOSAGE REGIMENS. Acta Obstet Gynecol Scand. 1977;56(S60):3–9. 12. Svanberg B, Arvidsson B, Norrby A, Rybo G, Sölvell L. ABSORPTION OF SUPPLEMENTAL IRON DURING PREGNANCY‐A LONGITUDINAL STUDY WITH REPEATED BONE‐MARROW STUDIES AND ABSORPTION MEASUREMENTS. Acta Obstet Gynecol Scand. 1975;54(48 S):87–108. 13. O’Brien KO, Ru Y. Iron status of North American pregnant women: An update on longitudinal data and gaps in knowledge from the United States and Canada. Am J Clin Nutr. 2017;106(Suppl 6):1647S-1654S. 14. Flegal KM, Graubard BI. Estimates of excess deaths associated with body mass index and other anthropometric variables. Am J Clin Nutr. 2009;89(4):1213–1219. 15. Kotelchuck M. An evaluation of the Kessner Adequacy of Prenatal Care Index and a 78 proposed Adequacy of Prenatal Care Utilization Index. Am J Public Health. 1994;84(9):1414–1420. 16. CDC, Satcher D, Hughes JM, et al. Recommendations to Prevent and Control Iron Deficiency in the United States. MMWR Recomm Rep. 1998;12(47(RR–3)):210. 17. Stevens GA, Finucane MM, De-Regil LM, et al. Global, regional, and national trends in haemoglobin concentration and prevalence of total and severe anaemia in children and pregnant and non-pregnant women for 1995-2011: A systematic analysis of population- representative data. Lancet Glob Heal. 2013;1(1):e16–e25. 18. Ohuma EO, Young MF, Martorell R, et al. International values for haemoglobin distributions in healthy pregnant women. EClinicalMedicine. 2020;29–30:. 19. Georgieff MK, Krebs NF, Cusick SE. The Benefits and Risks of Iron Supplementation in Pregnancy and Childhood. Annu Rev Nutr. 2019;39:121–146. 20. Mohamed MA, Ahmad T, MacRi C, Aly H. Racial disparities in maternal hemoglobin concentrations and pregnancy outcomes. J Perinat Med. 2012;40(2):141–149. 21. Adams PC, Reboussin DM, Barton JC, et al. Hemochromatosis and iron-overload screening in a racially diverse population. N Engl J Med. 2005; 22. Gordeuk VR, Brannon PM. Ethnic and genetic factors of iron status in women of reproductive age. Am J Clin Nutr. 2017;106:1594S-1599S. 23. Ye K, Cao C, Lin X, O’Brien KO, Gu Z. Natural selection on HFE in Asian populations contributes to enhanced non-heme iron absorption. BMC Genet. 2015;16(1):61. 79 CHAPTER 3 IRON SUPPLEMENTATION IN ANEMIC ZANZIBARI TODDLERS IS ASSOCIATED WITH GREATER LOSS IN ERYTHROCYTE IRON ISOTOPE ENRICHMENT * * Wanhui Kang, Nathaniel Baer, Mahdi Ramsan, Francoise Vermeylen, Rebecca Stoltzfus, Kimberly O'Brien, Iron supplementation in anemic Zanzibari toddlers is associated with greater loss in erythrocyte iron isotope enrichment. The American Journal of Clinical Nutrition, Volume 114, Issue 1, July 2021, Pages 330–337, https://doi.org/10.1093/ajcn/nqab044. 80 Abstract Background: Heavy parasitic loads increase the risk of iron deficiency anemia, which remains prevalent globally. Where parasites are common, understanding the influence of parasitic infections on Fe incorporation and erythropoiesis in toddlers is especially important. Objective: The aim of this study was to identify the impact of malarial and helminth infections on red blood cell (RBC) Fe incorporation and subsequent change in RBC Fe isotope enrichment for 84 days post-dosing in toddlers at high risk for parasitic infections. Methods: Fe incorporation was measured in a group of Zanzibari toddlers (n = 71, 16-25 months) using a stable Fe isotopic method. At study entry, oral stable Fe isotope was administered. Blood was collected 14 (D14) and 84 (D84) days post-dosing for assessment of Fe status indicators and RBC isotopic enrichment. Blood and stool samples were collected and screened for malaria and helminth parasites. Factors associated with change in RBC Fe isotope enrichment were identified using regression models. Results: Toddlers who had larger weight-for-age z-scores, lower total body Fe and those with helminth infections (n = 26) exhibited higher RBC Fe incorporation. RBC Fe isotope enrichment decreased from D14 to D84 by -2.75 percentage points (P < 0.0001, n = 66). Greater loss in RBC Fe isotope enrichment from D14 to D84 was observed in those who received Fe supplementation, those with either helminths or both malarial and helminth infections, and in those with greater RBC Fe incorporation on D14. Conclusion: Toddlers who received iron supplementation exhibited significantly greater loss of RBC iron isotope enrichment over time. We speculate this greater loss of RBC iron enrichment 81 is indicative of increased erythropoiesis due to the provision of iron among anemic or helminth infected toddlers. 82 Introduction Iron deficiency and iron deficiency anemia (IDA) are particularly detrimental during the perinatal period and low iron stores at this time have been associated with long-term cognitive and developmental deficits (1). In populations with IDA, iron availability is not sufficient to meet erythropoietic demands. Toddlers living in helminth- and malaria-endemic areas frequently suffer from chronic parasitemia and low-grade inflammation. These parasitic infections coupled with low dietary iron intakes increase the risk of IDA in many developing countries. Mechanisms by which malaria and helminth infection impair iron homeostasis are multifactorial. Malaria may exacerbate iron deficiency by sequestrating iron into malarial pigment (hemozoin) (2) and by blocking iron recycling from senescent red blood cells (RBC). It may also shorten the RBC lifespan due to erythrophagocytosis of both parasitized and non-parasitized RBC’s (3,4). Helminth infections can lead to chronic gastrointestinal blood loss increasing the risk of iron deficiency and anemia (5). In addition, inflammation caused by malarial and helminth infections may further compromise iron absorption due to increased serum hepcidin concentrations (3,4). Iron supplementation can be used to treat iron deficiency and IDA, but concerns exist when providing supplemental iron to children residing in malaria-endemic areas (6,7), particularly if they are iron replete (8). In malaria-endemic areas, non-targeted iron supplementation has been associated with increased mortality of preschool children (7,9). For this reason, the WHO now recommends screening for iron deficiency before iron supplementation is initiated in malaria-endemic locations (10). In addition, iron supplementation combined with anti-malarial medication is commonly used as an effective strategy for managing 83 post-malarial anemia (11). However, the mechanisms of how helminth and malarial infections interact to impact iron absorption and response to treatment are not fully understood (12). A better understanding of the differences and cumulative impact of these infections on iron metabolism in children is needed. This study was undertaken on Pemba Island, Zanzibar in 2004 when the overall prevalence of Plasmodium falciparum malaria was estimated to be 80% with no distinct evidence of seasonality (13,14). Moreover, among those aged 10-11 months, helminth and hookworm infections on Pemba island were endemic (13). The primary outcomes of this study were RBC iron isotope incorporation at D14 and changes in RBC iron enrichment over an 84-day period post dosing. The current study was undertaken to a) identify the impact of malarial and helminth infections on iron incorporation into RBC’s in relation to iron status indicators, and b) to identify variables associated with the change in RBC iron isotope enrichment over an 84-day period post- dosing among a group of toddlers at high risk for malarial and helminth infections. Subjects and Methods Subject recruitment and sample collection Participants were recruited to represent a random sample of healthy toddlers between 16 and 25 months of age with no other known health conditions. Toddlers that had fevers, symptomatic malaria, sickle cell trait or severe anemia (hemoglobin (Hb) < 7 g/dL) were not recruited into the study. A total of 82 toddlers were recruited from local villages on Pemba Island in November 2004. The stable isotope dosing study was initiated in December 2004 and ended in 84 March 2005. The study protocol was approved by the Committee of Human Research at the Johns Hopkins Bloomberg School of Public Health and the Ministry of Health, Zanzibar. Informed verbal consent was obtained from families willing to participate in the study. On the first day of the study parents were asked about their children’s diet and recent illnesses. Each toddler received a physical examination and anthropometric z-scores were calculated using the WHO database (15). A finger prick was obtained to assess Day 0 (D0) Hb, and malaria parasitic count. After D0 measurements were obtained, toddlers ingested a meal of porridge and remained fasted for 1.5 hours before ingesting a stable iron isotope as ferrous sulfate. Each participant received either 58Fe (0.9 mg of 58Fe, along with 6.0 mg of native ferrous sulfate to obtain a total iron dose of 7.0 mg, n = 37) or 57Fe (7.0 mg of 57Fe, n = 34) based on isotope availability. Tracer doses were administered with flavored raspberry syrup containing 0.391 % ascorbic acid (Humco, Texarkana, TX). Toddlers remained fasted for the next 1.5 hours before ingesting a second porridge meal. Two weeks post-dosing (D14), a 3 mL venous blood sample was obtained. Approximately three months post-dosing (D84), an additional 5 mL of blood was collected. Following the D84 clinic visit, all participants were treated with mebendazole as previously detailed (16). Malaria and helminths screening On D0, D14 and D84, whole blood was screened for malaria parasites using both thick and thin blood films (17). When the D14 blood was obtained, a stool sample was collected for analysis of helminth burden (Ascaris Lumbricoides, Trichuris Trichiura and Hookworm) using 85 the Kato-Katz method (18). Toddlers that had a positive blood smear for malarial parasites on D0, D14 or D84 were treated for malaria with artusonate/amodaquine. Hematological and iron status measures Hb concentrations were measured using a Hemocue (Lake Forest, CA). All iron status indicators were analyzed in both D14 and D84 blood samples, except soluble transferrin receptor (sTfR) which was only analyzed in the D14 samples. Any participant identified as anemic on D14 (Hb < 11.0 g/dL) received liquid iron supplementation (12.5 mg ferrous sulfate/day) (19), from D14 through D84. Serum ferritin (SF) and sTfR were measured by enzyme-linked immunosorbent assay (Ramco Laboratories, Inc Stafford, TX). Total body iron (TBI) was calculated by the ratio of sTfR to SF (TBI (mg/kg) = -[log sTfR*1000/SF) -2.8229]/0.1207) (20,21). Erythropoietin (EPO) and C-reactive protein (CRP) were measured using an Immulite® 1000 Analyzer (Seimens Healthcare Diagnostics, Deerfield, IL). Depleted iron stores were defined when SF concentrations were less than 12 ug/L and CRP concentrations were less than 5 mg/L, or when SF concentrations were less 30 ug/L and CRP concentrations were greater than 5mg/L. Elevated sTfR was classified if concentrations were greater than 8.5 mg/L. Elevated CRP was defined when concentrations were greater than 10 mg/L. Archived serum samples were used to measure hepcidin by a competitive enzyme-linked immunosorbent assay (Intrinsic Life Sciences, La Jolla, CA) on D14 serum (n = 42) and D84 serum (n = 38). Hepcidin values below the limit of detection (5 ng/mL) were assigned a concentration of 2.5 ng/mL. 86 RBC iron enrichment and incorporation Iron isotopic ratios were analyzed using magnetic sector thermal ionization mass spectrometry (ThermoScientific, Triton TI, Bremen, FRG). Mass spectrometric methods have been previously described in detail (22,23). The delta percent excess (∆%xs) of each administered stable iron isotope in RBC was determined by the observed (obs) and natural abundance (NA) isotope ratios: 57 Fe/ 56Fe 57 56 57 obs − Fe/ Fe NA ∆%xs Fe = × 100 57Fe/ 56Fe NA A similar equation was used for the 58Fe tracer. Natural abundance ratios used for the 57Fe/56Fe and 58Fe/56Fe ratios were 0.02317686 and 0.0030856, respectively. To account for the small differences in tracer doses administered, a Dose Adjusted Enrichment (DAE) was calculated for each tracer based on the mean dose of each isotope administered. For those receiving the 57Fe tracer, the mean dose of 57Fe administered was 6.64 mg 57Fe. The 57Fe DAE was calculated as: ∆%xs 57Fe × Average Dose of 57Fe 57Fe DAE = Actual Dose of 57Fe A similar equation was utilized to obtain a DAE after calculating the mean dose of 58Fe dose (0.891 mg 58Fe) that would give the same enrichment as that obtained with the average 57Fe dose administered. Throughout the manuscript the term RBC iron DAE reflects this value. In addition to the direct measurements of RBC iron DAE, RBC iron incorporation was calculated in D14 blood as previously reported (22,23). The blood volume in toddlers was estimated using the Linderkamp equation Log (Blood Volume) = 0.6459* (Log Weight in kg) + 0.002743 *(Height in cm) + 2.0324 (24). The change in RBC iron DAE from D14 to D84 post-dosing was determined by subtracting each toddler’s D14 DAE from the D84 DAE. Observed decreases in 87 RBC iron DAE over time may occur due to increases in the size of the circulating RBC pool induced by absorption of natural isotopic iron or mobilization of storage iron (25). Statistical analysis Descriptive statistics were used to summarize characteristics of participants, using the mean ± standard deviation (SD) or geometric mean (95% confidence interval (CI)) for continuous variables or frequency counts with percentages for categorical variables. Student's t tests and Wilcoxon’s rank-sum tests were used to assess possible differences in concentrations of biochemical indicators between groups stratified by sex, anemia, presence of malarial or helminth infections, or presence of both parasites. Pearson correlation or Spearman rank correlations were used to evaluate the strength of the bivariate linear association between variables. The Chi-squared test of independence was used for analyzing the association between categorical variables. The odds ratio of anemia was evaluated using logistic regression after adjustment for confounders (sex, age, and anthropometric measures). Multiple linear regression was used to estimate the association of RBC iron incorporation and change in RBC iron DAE after controlling for confounding factors such as sex and age. All analyses were performed using the JMP Pro statistical program version 13.1.0 (SAS Institute, Inc., Cary, NC). Significance was defined by a P value less than 0.05. 88 Results Characteristics of the study participants In total, 82 women consented to allow their child to participate in the study and among this group, 71 toddlers completed the study as detailed in Figure 3.1. Figure 3.1 Flow chart of toddlers enrolled in the study. Overview of the parasitic infections in Zanzibari toddlers aged 16-25 months. F, Females; M, Males. Characteristics of participants are reported in Table 3.1. More female (n = 41) than male toddlers (n = 30) enrolled in the study. There were no statistically significant differences in the percentage 89 of stunting or wasting between the males and females, but females tended to be more underweight than the males (P < 0.08). At enrollment on D0, 16% (11/70) of participants had a positive blood smear for malarial parasites. At D14, 37% (26/70) and 11% (8/70) of toddlers had helminths and malarial parasitemia, respectively. At D14, 6% (4/70) of toddlers were infected with both helminths and malaria. Prevalence of malaria and helminth infections did not significantly differ between males and females. On D14, 47% (n = 33) toddlers were anemic and were treated with iron supplementation (iron Supplemented Group). The remaining 53% of children were non-anemic and were not supplemented with iron (Non-iron supplemented Group). Characteristics of these 2 groups of toddlers are presented in Table 3.2. The D84 blood sample was collected 84 days after tracer dosing in all but one toddler in whom the blood sample was collected 83 days post-dosing. Table 3.1 General characteristics of the 71 toddlers enrolled in the study 1 All Toddlers Females Males (n = 71) (n = 41) (n = 30) Age, month 19.47 ± 2.19 19.41 ± 2.06 19.55 ± 2.39 Height, cm2 77.18 ± 4.23 75.91 ± 3.60 78.91 ± 4.48 Weight, kg3 9.78 ± 1.41 9.13 ± 0.97 10.66 ± 1.45 Weight-for-age z-scores4 -0.93 ± 1.06 -1.22 ± 0.93 -0.55 ± 1.13 Underweight, %5 15.49 [11] 21.95 [9] 6.67 [2] Length-for-age z-scores -1.91 ± 1.19 -2.06 ± 1.09 -1.71 ± 1.29 Stunted, % 39.43 [28] 41.46 [17] 36.67 [11] Weight-for-length z-scores 0.00 ± 1.28 -0.26 ± 0.98 0.36 ± 1.55 Wasted, % 2.82 [2] 2.44 [1] 3.33 [1] 1Data are presented as the mean ± SD or % [n]. Differences between sexes were analyzed by Student’s t test or Chi-square test. 2Significant difference observed between the males and females, P < 0.005. 90 3Significant difference observed between the males and females, P < 0.0001. 4Significant difference observed between the males and females, P < 0.01. 5Difference between males and females approached significance, P = 0.08. Table 3.2 Iron status in anemic toddlers that received, or did not receive, iron supplementation on D141 Time- Iron Supplemented Group Non-Iron Supplemented Group point D0 10.38 ± 1.41 [33] 11.19 ± 1.42 [35]2 Hb, g/dL D14 9.96 ± 1.02 [33] 12.27 ± 1.00 [36]3 D84 11.33 ± 1.30 [32]4 12.28 ± 1.09 [33]5 D14 27.76 (20.54, 37.51) [32] 30.98 ± 19.74 [36] SF, µg/L D84 34.42 (24.29, 48.78) [32]6 19.75 (14.77, 26.39) [30]2 sTfR, mg/L D14 9.03 (7.82, 10.43) [33] 7.76 ± 2.36 [36]2 D14 18.67 (13.66, 25.50) [31] 11.15 (8.76, 14.19) [36]2 EPO, IU/L D84 12.14 (9.18, 16.07) [30]7 11.49 (8.78, 15.05) [31] D14 0.27 (0.15, 0.49) [31] 0.32 (0.22, 0.48) [36] CRP, mg/L D84 0.37 (0.22, 0.62) [32] 0.28 (0.13, 0.58) [31] D14 0 0 Elevated CRP, % D84 3.13 [1/32] 3.13 [1/32] Serum hepcidin, D14 26.62 ± 21.12 [17] 14.04 (7.56, 26.08) [24] ng/mL D84 48.22 (26.10, 89.08) [16]7 23.64 (12.02, 46.49) [21] TBI, mg/kg D14 2.56 ± 3.09 [32] 2.90 ± 2.72 [36] 1Data are presented as the mean ± SD [n] or geometric mean (95% CI) [n] or % [n]. Elevated CRP was defined as concentrations greater than 10 mg/L. Differences between toddlers supplemented with iron and without iron supplementation were analyzed by Student’s t test or Wilcoxon’s rank-sum test. Differences between D14 and D84 measurements were analyzed by paired Student’s t test. CRP, C-reactive protein; D0, first day of the study visit; D14, 14 days post-dosing; D84, 84 days post-dosing; EPO, erythropoietin; Hb, hemoglobin; SF, serum ferritin; sTfR, soluble transferrin receptor; TBI, total body iron. 2Significant difference observed between toddlers supplemented with iron and without iron supplementation, P < 0.05. 3Significant difference observed between toddlers supplemented with iron and without iron supplementation, P < 0.005. 4Significant change from D14 to D84 observed in toddlers supplemented with iron supplementation, P < 0.0001. 91 5Significant difference observed between toddlers supplemented with iron and without iron supplementation, P < 0.0001. 6A tendency of significant change from D14 to D84 observed in toddlers supplemented with iron supplementation, P = 0.08. 7Significant change from D14 to D84 observed in toddlers supplemented with iron supplementation, P < 0.05. Hematological and iron status On D0, 56% (40/71) of toddlers were anemic. Those with elevated sTfR were nearly 5- fold more likely to be anemic (Odds Ratio: 4.69; 95% CI: 1.37, 13.98). Interestingly, toddlers who had helminth infections had a 40% lower risk of anemia when compared to those without helminth infections. After adjusting for sex, age and anthropometric measurements, anemia was significantly less prevalent in helminth infected toddlers (Adjusted odds ratio (AOR): 0.26; 95% CI: 0.07, 0.97) and significantly more prevalent among those with elevated sTfR (AOR: 4.74; 95% CI: 1.01, 22.15). As expected, mean Hb concentration increased from D14 to D84 in those who were iron supplemented, but no significant change in Hb was observed in the non-iron supplemented group (P = 0.65) (Table 3.2). In addition, iron supplemented toddlers had a significant decrease in EPO from D14 to D84 as well as a tendency for SF concentrations to increase (Table 3.2). There were no significant changes in EPO or SF in the non-iron supplemented group. Hepcidin analyses were only conducted in a sub-group of toddlers (n = 41 at D14 and n = 37 at D84) based on availability of serum. Hepcidin concentrations on both D14 and D84 did not significantly differ between the anemic and non-anemic participants. Hepcidin concentrations, however, did increase significantly from D14 to D84 in the iron supplemented group but not among the non-iron supplemented group (P = 0.32) (Table 3.2). At D14, hepcidin was inversely 92 correlated with D14 sTfR (P < 0.05, r = -0.42) and positively correlated with D14 SF (P < 0.05, r = 0.34). There was a lack of correlation between hepcidin and CRP on D14. At D84, hepcidin concentrations remained significantly correlated with D84 SF concentrations (P = 0.005, r = 0.42) and were also positively correlated with D84 CRP values (P < 0.05, r = 0.49). Participants with malarial parasitemia on D0 had impaired iron and hematological status as evident by significantly lower D14 TBI, higher D14 sTfR and D14 EPO concentrations and a tendency for lower D0 Hb concentrations (Table 3.3). The iron status of toddlers with malarial parasitemia on D14 is presented in Table 3.3. The iron status of those with helminths did not differ from those without helminths at D14. For the subset of toddlers who had hepcidin measured, D14 hepcidin was significantly lower in those who had helminth infections (P = 0.03, n = 42) (data not shown). 93 Table 3.3 Iron status as a function of malaria parasitemia1 Time Malaria Parasitemia at D0 Malaria Parasitemia at D14 - point Yes No Yes No 9.98 ± 1.46 11.00 ± 1.42 10.19 ± 1.74 D0 10.91 ± 1.42 [61] [11] [58]2 [8] 11.00 ± 1.00 11.17 ± 1.60 10.74 ± 2.07 11.23 ± 1.45 Hb, g/dL D14 [11] [58] [8] [62] 12.20 ± 0.91 11.70 ± 1.32 11.66 ± 0.92 11.85 ± 1.33 D84 [11] [55] [8] [58] 21.5 29.3 41.4 26.3 D14 (14.4, 32.3) (23.0, 37.5) (28.1, 61.1) (20.9, 33.2) [61]3 [11] [57] [8] SF, µg/L 20.9 28.4 24.7 26.6 D84 (11.7, 37.5) (22.1, 36.6) (13.4, 45.3) (20.7, 34.5) [55] [11] [52] [8] 10.49 7.77 9.28 sTfR, 7.97 D14 (9.12, 12.06) (7.03, 8.55) (7.27, 11.85) mg/L (7.26, 8.76) [62] [11] [58]4 [8] 21.97 13.20 25.51 13.20 D14 (13.90, 34.71) (10.59, 16.34) (13.94, 46.68) (10.70, 16.20) [61]3 EPO, [11] [56]4 [7] IU/L 17.77 10.99 16.09 11.25 D84 (11.72, 26.96) (8.94, 13.46) (8.86, 29.25) (9.21, 13.74) [54] [11] [51]4 [8] 0.28 0.87 0.44 0.28 D14 (0.19, 0.41) (0.27, 2.83) (0.20, 0.97) [11] (0.19, 0.39) [61] CRP, [56] [7] mg/L 0.32 1.55 0.35 0.25 D84 (0.20, 0.51) (0.39, 6.23) (0.11, 1.17) [11] (0.16, 0.39) [56]3 [53] [8] Elevated D14 0 0 0 0 CRP, % D84 9.09 [1/11] 1.89 [1/53] 12.5 [1/8] 1.79 [1/56] 17.37 11.88 24.85 (16.36, 37.75) 14.15 (8.61, 23.18) D14 (10.58, 28.49) Serum (4.31, 32.75) [9] [6] [36] [32] hepcidin, 44.70 53.25 ng/mL 14.95 29.07 (18.23,46.35) D84 (27.54, 72.68) (8.69, 327.01) (5.93, 37.67) [9] [32] [28]4 [6] TBI, 3.52 {4.16} 3.28 {2.64} D14 1.12 ± 2.54 [11] 3.37 {4.13} [61] mg/kg [57]4 [8] 94 1Data are presented as the mean ± SD [n] or geometric mean (95% CI) [n] or median {interquartile range} [n] or % [n]. Elevated CRP was defined as concentrations greater than 10 mg/L. Differences between toddlers with and without malaria parasitemia at D0 or D14 were analyzed by Student’s t test or Wilcoxon’s rank-sum test. CRP, C-reactive protein; D0, first day of the study visit; D14, 14 days post- dosing; D84, 84 days post-dosing; EPO, erythropoietin; Hb, hemoglobin; SF, serum ferritin; sTfR, soluble transferrin receptor; TBI, total body iron. 2Difference between toddlers with and without malaria parasitemia at D0 approached significance, P = 0.05. 3Significant difference observed between toddlers with and without malaria parasitemia at D14, P < 0.05. 4Significant difference observed between toddlers with and without malaria parasitemia at D0, P < 0.05. RBC iron incorporation Iron incorporation data are presented in Table 3.4. RBC iron incorporation at D14 was highly variable (ranging from 1.21 - 55.45%). As expected, average RBC iron incorporation did not significantly differ between those that received the 58Fe or the 57Fe tracer. TBI was correlated with RBC iron incorporation (P = 0.009, r = -0.32, n = 69) but there were no other significant associations with other iron status indicators or with hepcidin. Infection with both parasites was associated with significantly higher RBC iron incorporation, whereas infection with helminths only tended to be associated with higher RBC iron incorporation (P = 0.07). Toddlers with malarial parasitemia did not have higher RBC iron incorporation compared to those without this infection (Table 3.4). The association between iron incorporation and parasitic infections did not appear to be driven by low iron status as there were no significant associations between presence of either or both parasitic infections with any of the iron status indicators measured. Mean D14 hepcidin tended to be higher (P = 0.07) in those without malaria and/or helminth infections compared to the mean observed in those with either or both of these parasitic infections (10.24 ng/mL). In the subgroup of participants with either malarial or helminth infection, RBC iron incorporation was negatively correlated with CRP (P < 0.05, r = -0.38, n = 95 31) but was not significantly impacted by hepcidin on D14. RBC iron incorporation did not differ between the iron supplemented or non-supplemented toddlers (P = 0.5). Significantly higher RBC iron incorporation was evident among those with higher weight-for-age z-scores, lower TBI, and among those with helminth infections (P < 0.05). Hepcidin was not significantly associated with RBC iron incorporation, which may be due to the high variability in hepcidin coupled with the known impact of infection on the concentrations of this biomarker. Table 3.4 Factors associated with red blood cell iron incorporation (%) at D141 Groups Iron incorporation at D14 (%) P All toddlers 11.42 (9.29, 14.02) [68] --- Females 10.89 (8.44,14.05) [40] NS Males 12.22 (8.52, 17.53) [28] D14 Anemia 11.41 (8.54, 15.25) [33] NS No D14 Anemia 11.42 (8.41, 15.52) [35] D14 Helminths 14.46 (10.54, 19.82) [26] 0.07 No D14 Helminths 9.87 (7.53, 12.92) [42] D14 Malaria parasitemia 10.03 (4.44, 22.68) [8] NS No D14 Malaria parasitemia 11.61 (9.35, 14.43) [60] Presence of parasites 14.50 (10.97, 19.15) [31] 0.03 No parasites 9.17 (6.80, 12.37) [36] D14 Detectable hepcidin 8.55 (6.16,11.87) [28] 0.08 D14 Undetectable hepcidin 14.76 (8.63, 25.27) [12] 1Data are presented as the geometric mean (95% CI) [n]. Comparison groups are stratified by sex, presence of anemia on D14, presence of malarial or helminth infections on D14 or either parasites over the study period, or by hepcidin concentrations on D14. Hepcidin values below 5 ng/mL were defined as undetectable hepcidin. Differences were analyzed by Student’s t test or Wilcoxon’s rank-sum test. D14, 14 days post-dosing; NS, not significant. 96 Change in RBC iron DAE from D14 to D84 RBC iron DAE decreased from D14 to D84 on average by 2.75 percentage-points (from 15.94% to 13.19%; 95% CI: -3.86 percentage-points, -1.65 percentage-points, P < 0.0001, n = 66), but the degree of change varied more than 6-fold from -16.41 percentage-points to +12.75 percentage-points. Among the 66 toddlers with data available on the change in RBC iron DAE between D14 and D84, 12 toddlers (28%) exhibited slight increases in RBC iron DAE from D14 to D84. Those with increases in RBC iron DAE were equally distributed among toddlers that had received the 57Fe (n = 5) or 58Fe (n = 7) tracer. On average, in these 12 individuals, the mean increase observed was 1.67 ± 3.52 percentage points. After excluding the toddler with the highest increase (+12.75 percentage points), the mean RBC iron DAE increase observed was 0.66 ± 0.49 percentage points. Possible differences in those that exhibited increases in DAE over time versus those that exhibited decreases in DAE over time were explored. Toddlers that had undetectable hepcidin on D84 (P = 0.02) and toddlers with lower sTfR concentrations on D14 (P = 0.04) were significantly more likely to have had an increase in RBC iron DAE between D14 to D84. In addition, participants with lower RBC iron DAE at D14 were more likely (P = 0.04) to experience increased RBC iron DAE at D84. These results remained significant even when data from the toddler with the largest 12.75 percentage points increase were excluded. The net change in RBC iron DAE over time was not significantly impacted by sex, age, anthropometric measurements or malarial infections. A comparison of the change in RBC iron DAE between different groups is presented in Figure 3.2. 97 Figure 3.2 Mean change in RBC iron DAE (D84-D14) in toddlers as a function of sex, iron status or parasitic burden. Comparison groups are stratified by sex, iron supplementation on D14, presence of malarial or helminth infections on D14 or either parasites over the study period, or by hepcidin concentrations on D84. Hepcidin values below 5 ng/mL were defined as undetectable hepcidin. Number on each bar indicates the number of toddlers in each group. Differences were analyzed between each pair of groups by Wilcoxon’s rank-sum test, *P < 0.05, **P < 0.005. Error bars represent SEMs. DAE, Dose Adjusted Enrichment; D14, 14 days post- dosing; D84, 84 days post-dosing; RBC, red blood cell. Change in RBC iron DAE was significantly greater in those who were supplemented with iron (P = 0.001) versus those who had not been iron supplemented. This change in DAE in those receiving iron supplementation appeared to be due to the addition of iron and not a consequence 98 of the pre-existing anemia in this group, based on the finding that toddlers with anemia at D0 did not have a significantly greater change in RBC iron DAE compared to non-anemic toddlers at D0 (data not shown). Participants with higher hepcidin concentrations on D84 had greater reductions in RBC iron DAE compared to those with undetectable hepcidin at D84 (P = 0.003, n = 38). RBC iron DAE decreased to a larger extent in those found to be infected with either malaria or helminths (P = 0.02) across the study period. Toddlers with helminths on D14 had greater losses in RBC iron DAE over the 70-day observation period (P = 0.05) but the presence of malarial infections did not have a significant impact on change in RBC iron DAE over time (P = 0.19) perhaps due to the limited sample size of those with malaria (n = 8). Among all hematological and iron status indicators analyzed, Hb (either at D0 or D14) was the only biomarker that was significantly negatively associated with change in RBC iron DAE over time (P < 0.005). In addition, an increase in Hb from D14 to D84 was significantly correlated with greater decrease in RBC iron DAE from D14 to D84 (P < 0.005, r = 0.37). A multiple linear regression model was constructed to identify factors significantly associated with the change in RBC iron DAE. On D14, iron supplementation (β ± SE (standard error) = 0.40 ± 0.47; P = 0.0005), RBC iron incorporation (β ± SE = 0.44 ± 0.04; P = 0.0006) and helminth infection (β ± SE = 1.36 ± 0.54; P = 0.03) were found to be significantly associated with a decrease in RBC iron DAE from D14 to D84 (Adjusted R2 = 0.38; n = 62). A second model was constructed in the sub-group of toddlers with hepcidin data (Adjusted R2 = 0.54; n = 36) setting the change in RBC iron DAE as the dependent variable and iron incorporation and iron status indicators as independent variables. In this smaller sub-population, helminth infection was no longer significantly associated with change in RBC iron DAE but those with detectable 99 hepcidin at D84 exhibited a significantly greater loss in RBC iron DAE (β ± SE = 0.30 ± 0.62; P = 0.03). Discussion A better understanding of the impact of parasitic infections on iron metabolism is needed, especially if iron supplementation may exacerbate adverse events when given to iron replete children with active parasitic infections. To our knowledge, this study is one of the largest iron stable isotope studies to follow RBC iron DAE in this age group over an 84-day period. Toddlers with better weight-for-age z-scores, lower TBI and those with helminth infections exhibited higher RBC iron incorporation. A greater decrease in RBC iron DAE post-dosing was evident among those receiving iron and among those with helminth infections. An unexpected finding was that toddlers who had helminth infections had a nearly 40% lower risk of anemia compared to those without helminth infections. An earlier study in a larger cohort of Zanzibari children in the parent study (13) also found a lower prevalence of anemia in children (6-23 months) with helminth infections compared to children without helminth infections. In contrast, in older Zanzibari children (7-13 years), Stoltzfus et al. (26) found that hookworm infection was the strongest risk factor for severe anemia. Age differences between studies may in part explain these differences as anemic and poorly nourished toddlers may not be actively exploring their environments and thus may have less exposure to helminth-infected soil. Toddlers that experienced either malarial or helminth infection across the study had significantly higher RBC iron incorporation than those who remained parasite-free, an effect that appeared to be driven primarily by the non-significant increase in RBC iron incorporation in those with helminth infections. RBC iron incorporation did not differ between toddlers with and without malaria parasitemia but the small number with malaria parasitemia limited our ability to 100 address the independent effect of malaria on study outcomes. Participants in our study incorporated an average of 11.4% of ingested iron into their RBCs. This is comparable to the 9.0% and 11.5% RBC iron incorporation found among 48 Malawian toddlers (12–24 months) that received iron supplements (30 mg/day) either immediately, or two weeks after, being treated for malaria (27). After 2 weeks of oral iron supplementation, mean SF concentration increased by 2.5-fold and iron incorporation decreased significantly to 5.9% in the Malawian cohort (27). This result is consistent with our finding of a lower RBC iron incorporation in those with better iron status. Average RBC iron incorporation in our study was slightly lower than the 18% reported in 27 Gambian toddlers (18–36 months) who were dosed with a 43% lower quantity of iron tracer (4 mg) (21). Among all biomarkers evaluated, the strongest determinant of RBC iron incorporation was TBI. Our finding that lower iron status, but not decreased inflammation, explained more variability in RBC iron incorporation is in agreement with prior data (28). It is likely that TBI better captures iron demands by accounting for both tissue demand and storage iron, and TBI may attenuate the impact of inflammation in populations at risk for recurrent parasitic infections. Data from 156 Ugandan children (1.5–5 years) found that subclinical inflammation (evidenced by elevated hepcidin and CRP) persisted for 12 months after antimalarial treatment (11,29). In our cohort, hepcidin was not a significant determinant of RBC iron incorporation. This finding differs from that observed in a population of anemic Gambian toddlers (21), where hepcidin was found to be the best predictor of RBC iron incorporation. When toddlers are actively inflamed, hepcidin may explain more variability in RBC iron incorporation whereas iron status may be a better predictor of RBC iron incorporation in the absence of inflammation. We were unable to obtain measures of hepcidin on D0. It is likely that in this population with frequent infections, 101 the D14 hepcidin measure may not have adequately captured hepcidin status at the time tracer was administered and absorbed. Our data also indicate that helminth infections are associated with higher iron incorporation, which has also been reported in Ivorian children (30). This study uniquely followed the change in RBC iron isotopic enrichment over 84-days in healthy toddlers residing in a malaria- and helminth-endemic area. On average RBC iron DAE decreased over the 84-days, but a small subset of our cohort exhibited a slight increase in RBC iron DAE over time. Those that exhibited increases in RBC iron DAE were significantly more likely to have undetectable hepcidin at D84, and lower sTfR concentrations and RBC iron DAE at D14. The slight increase in RBC iron DAE observed draws attention to the need to further refine the timing used to evaluate RBC iron incorporation in healthy toddlers and/or in those with parasitic infections. This methodology and the measure of RBC incorporation 2-week after dosing was developed from early radiotracer data in adults (31). There are no other data evaluating changes in RBC iron enrichment over time in toddlers to determine if our findings are unique to our cohort, or if the timing of RBC iron incorporation is more variable in this age group. Thus, further normative data on this methodology in children are warranted. In these toddlers, decreases in RBC iron DAE were most pronounced in those who received iron supplementation and exhibited significant increases in Hb post-supplementation. Greater losses in RBC iron DAE over time did not appear to be driven by pre-existing anemia as the D14 to D84 decline in RBC iron DAE was not significantly impacted by the presence of anemia at D0. In addition, iron supplementation was a stronger determinant of change in RBC iron DAE than was the presence of malaria and/or helminth parasites. Two other studies in toddlers followed RBC iron enrichment over an extended period. Fomon et al. (32) studied 35 Caucasian infants (5.6 months) and found the change in RBC iron enrichment over 13 months 102 was significantly correlated with iron absorption, which is similar to our finding of greater decreases in RBC iron DAE among toddlers with higher RBC iron incorporation. More recently, Speich et al. (25) dosed a group of African toddlers (14-20 months) and allowed for isotopic equilibration to occur for an 8-11 month period post-dosing before evaluating the impact of iron supplementation on RBC tracer enrichment. Our study cannot be directly compared to the Speich study as we did not allow sufficient time for tracer to equilibrate with body iron pools but there are some similarities in findings. Consistent with their study, we found that hepcidin was not significantly correlated with RBC iron incorporation and observed decreases in RBC iron enrichment were positively correlated with iron absorption, increases in Hb and iron supplementation. They uniquely found that although iron supplementation increased the absolute amount of iron absorbed the additional iron absorbed did not appear to be retained (25). Our study contributes to the data available on factors that impact RBC iron incorporation and change in RBC iron DAE over time in toddlers residing in a malaria- and helminth-endemic area but there are several limitations. Inflammatory status was evaluated using CRP, but this may not be a robust inflammatory marker in those with chronic infections (33–35). We could not differentiate a new onset of malarial infection from a relapse of an existing infection. Although anti-malarial/helminths medications were provided, recurrence of new infections was difficult to prevent in a parasite-endemic context. Our classification of infection as one that involved either malaria or helminths likely increased our power to identify effects of these two parasitic infections, we could not fully evaluate the impact of each individual infection due in particular to the lower number of toddlers with malaria. In summary, RBC iron incorporation was best predicted by weight-for-age z-scores and TBI in these Zanzibari toddlers. RBC iron DAE decreased significantly across an interval of time 103 that approaches the typical RBC lifespan in this age group (36). Toddlers who received iron, incorporated more iron into RBC’s exhibited significantly greater losses in RBC iron DAE post- dosing. Future longer-term labeling studies may provide greater insight into whole body iron homeostasis and adaptations to parasitic loads or iron supplementation in these vulnerable age groups. 104 REFERENCES 1. Lozoff B, Beard J, Connor J, et al. Long-lasting neural and behavioral effects of iron deficiency in infancy. Nutr Rev 2006;64(5 Pt 2):S34-43; discussion S72-91. 2. Olivier M, Van Den Ham K, Shio MT, Kassa FA, Fougeray S. Malarial pigment hemozoin and the innate inflammatory response. Front Immunol 2014;5:25. 3. Cherayil BJ. Pathophysiology of iron homeostasis during inflammatory states. J Pediatr 2015;167(4 Suppl):S15-9. 4. Friedman JF, Kanzaria HK, McGarvey ST. Human schistosomiasis and anemia: The relationship and potential mechanisms. Trends Parasitol 2005;21(8):386-92. 5. Crompton DWT, Nesheim MC. Nutritional impact of intestinal helminthiasis during the human life cycle. Annu. Rev. Nutr. 2002;22:35–59. 6. Barffour MA, Schulze KJ, Coles CL, et al. High Iron Stores in the Low Malaria Season Increase Malaria Risk in the High Transmission Season in a Prospective Cohort of Rural Zambian Children. J Nutr 2017;147(8):1531-1536. 7. Spottiswoode N, Duffy PE, Drakesmith H. Iron, anemia and hepcidin in malaria. Front Pharmacol 2014;5:125. 8. Neuberger A, Okebe J, Yahav D, Paul M. Oral iron supplements for children in malaria- endemic areas. Cochrane Database Syst Rev 2016;2(2):CS006589. 9. Sazawal S, Black RE, Ramsan M, et al. Effects of routine prophylactic supplementation with iron and folic acid on admission to hospital and mortality in preschool children in a high malaria transmission setting: Community-based, randomised, placebo-controlled trial. Lancet 2006;367(9505):133-43. 105 10. Cusick SE, Georgieff MK, Rao R. Approaches for reducing the risk of early-life iron deficiency-induced brain dysfunction in children. Nutrients 2018;10(2):227. 11. Cusick SE, Opoka RO, Ssemata AS, Georgieff MK, John CC. Delayed iron improves iron status without altering malaria risk in severe malarial anemia. Am J Clin Nutr 2020; 111(5):1059-1067. 12. Doherty CP, Cox SE, Fulford AJ, et al. Iron incorporation and post-malaria anaemia. PLoS One 2008;3(5):e2133. 13. Kung’u JK, Goodman D, Haji HJ, et al. Early helminth infections are inversely related to anemia, malnutrition, and malaria and are not associated with inflammation in 6- to 23- month-old Zanzibari children. Am J Trop Med Hyg 2009;81(6):1062-70. 14. Jaenisch T, Sullivan DJ, Dutta A, et al. Malaria incidence and prevalence on Pemba Island before the onset of the successful control intervention on the Zanzibar Archipelago. Malar J 2010; 9:32. 15. World Health Organization, Multicentre Growth Reference Study Group. WHO Child Growth Standards: Length/height-for-age, weight-for-age, weight-for-length, weight-for- height and body mass index-for-age: Methods and development. Geneva, Switzerland: World Health Organization, 2006. 16. Stoltzfus RJ, Albonico M, Chwaya HM, et al. Effects of the Zanzibar school-based deworming program on iron status of children. Am J Clin Nutr 1998;68(1):179-86. 17. Moody A. Rapid diagnostic tests for malaria parasites. Clin Microbiol Rev 2002;15(1):66-78. 18. Martin LK, Beaver PC. Evaluation of Kato thick-smear technique for quantitative diagnosis of helminth infections. Am J Trop Med Hyg 1968;17(3):382–391. 106 19. World Health Organization. Guideline: Daily iron supplementation in infants and children. Geneva, Switzerland: World Health Organization, 2016. 20. Cook JD, Flowers CH, Skikne BS. The quantitative assessment of body iron. Blood 2003; 101 (9): 3359–3363. 21. Prentice AM, Doherty CP, Abrams SA, et al. Hepcidin is the major predictor of erythrocyte iron incorporation in anemic African children. Blood 2012;119(8):1922–8. 22. Abrams SA, Wen J, O’Brien KO, Stuff JE, Liang LK. Application of magnetic sector thermal ionization mass spectrometry to studies of erythrocyte iron incorporation in small children. Biol Mass Spectrom 1994;23(12):771-5. 23. Abrams SA, O’Brien KO, Wen J, Liang LK, Stuff JE. Absorption by 1-year-old children of an iron supplement given with cow’s milk or juice. Pediatr Res 1996;39(1):171-5. 24. Linderkamp O, Versmold HT, Riegel KP, Betke K. Estimation and prediction of blood volume in infants and children. Eur J Pediatr 1977;125(4):227–234. 25. Speich C, Wegmüller R, Brittenham GM, et al. Measurement of long‐term iron absorption and loss during iron supplementation using a stable isotope of iron (57 Fe). Br J Haematol 2020 Aug 30 (Epub ahead of print; DOI: 10.1111/bjh.17039). 26. Stoltzfus RJ, Chwaya HM, Tielsch JM, et al. Epidemiology of iron deficiency anemia in Zanzibari schoolchildren: the importance of hookworms. Am J Clin Nutr 1997;65(1):153–159. 27. Glinz D, Kamiyango M, Phiri KS, et al. The effect of timing of iron supplementation on iron absorption and haemoglobin in post-malaria anaemia: a longitudinal stable isotope study in Malawian toddlers. Malar J 2014;13:397. 107 28. Cusick SE, Opoka RO, Abrams SA, et al. Delaying Iron Therapy until 28 Days after Antimalarial Treatment Is Associated with Greater Iron Incorporation and Equivalent Hematologic Recovery after 56 Days in Children: A Randomized Controlled Trial. J Nutr 2016;146(9):1769–74. 29. Suchdev PS, Trehan I. Optimizing iron supplementation for children with severe malaria. Am J Clin Nutr 2020;111(5):939-940. 30. Glinz D, Hurrell RF, Righetti AA, et al. In Ivorian school-age children, infection with hookworm does not reduce dietary iron absorption or systemic iron utilization, whereas afebrile Plasmodium falciparum infection reduces iron absorption by half. Am J Clin Nutr 2015;101(3):462-70. 31. Bothwell TH. Iron metabolism in man. Oxford: Blackwell Scientific Publications; 1979:175-6. 32. Fomon SJ, Nelson SE, Serfass RE, Ziegler EE. Absorption and Loss of Iron in Toddlers Are Highly Correlated. J Nutr 2005;135(4):771–777. 33. Suchdev PS, Williams AM, Mei Z, et al. Assessment of iron status in settings of inflammation: Challenges and potential approaches. Am J Clin Nutr 2017;106(Suppl 6):1626S-1633S. 34. Raiten DJ, Sakr Ashour FA, Ross AC, et al. Inflammation and nutritional science for programs/policies and interpretation of research evidence (INSPIRE). J Nutr 2015;145(5):1039S-1108S. 35. Namaste SML, Ou J, Williams AM, et al. Adjusting iron and vitamin A status in settings of inflammation: a sensitivity analysis of the Biomarkers Reflecting Inflammation and 108 Nutritional Determinants of Anemia (BRINDA) approach. Am J Clin Nutr 2020;112(1):458S-467S. 36. Fomon SJ, Serfass RE, Nelson SE, Rogers RR, Frantz JA. Time Course of and Effect of Dietary Iron Level on Iron Incorporation into Erythrocytes by Infants. J Nutr 2000;130(3):541–545. 109 CHAPTER 4 ETHNIC DIFFERENCES IN IRON STATUS* * Wanhui Kang, Alexa Barad, Andrew G. Clark, Yiqin Wang, Zhenglong Gu, Kimberly O. O’Brien. Ethnic differences in iron status. Advances in Nutrition, 2021;00:1–16; doi:https://doi.org/10.1093/advances/nmab035. 110 Abstract Iron is unique among all minerals in that humans have no regulatable excretory pathway to eliminate excess iron after it is absorbed. Iron deficiency anemia occurs when absorbed iron is not sufficient to meet body iron demands, whereas iron overload and subsequent deposition of iron in key organs occurs when absorbed iron exceeds body iron demands. Over time, iron accumulation in the body can increase risk of chronic diseases, including cirrhosis, diabetes, and heart failure. To date, only ~30% of the interindividual variability in iron absorption can be captured by iron status biomarkers or iron regulatory hormones. Much of the regulation of iron absorption may be under genetic control, but these pathways have yet to be fully elucidated. Genome-wide and candidate gene association studies have identified several genetic variants that are associated with variations in iron status, but the majority of these data were generated in European populations. The purpose of this review is to summarize genetic variants that have been associated with alterations in iron status and to highlight the influence of ethnicity on the risk of iron deficiency or overload. Using extant data in the literature, linear mixed-effects models were constructed to explore ethnic differences in iron status biomarkers. This approach found East Asians had significantly higher concentrations of iron status indicators (serum ferritin, transferrin saturation and hemoglobin) compared to Europeans, African Americans, or South Asians. African Americans exhibited significantly lower hemoglobin concentrations compared to other ethnic groups. Further studies of the genetic basis for ethnic differences in iron metabolism and on how it affects disease susceptibility among different ethnic groups are needed to inform population-specific recommendations and personalized nutrition interventions for iron-related disorders. 111 Introduction Iron is an essential trace element involved in numerous metabolic processes, including oxygen transport and utilization, cellular proliferation, DNA synthesis, neurotransmitter synthesis, and energy production (1,2). Iron is the fourth most common element in the Earth’s crust (1,3), yet iron deficiency (ID) remains prevalent, affecting an estimated 2.5 billion people worldwide (4). Iron overload (IO), on the other hand, is associated with adverse health outcomes caused by the accumulation of iron in organs particularly the pancreas, liver and heart. Elevated iron stores are associated with increased risk of type 2 diabetes independent of inflammation (5,6), and increased risk of cardiovascular disease (7,8), liver fibrosis, and cancer (9–11). Moreover, age-related iron accumulation in the brain strongly predicts cognitive decline, motor impairment, and the development of neurodegenerative diseases (12,13). A causal role of enhanced iron status in these diseases has been noted in animal models of IO (14,15). Increased attention has been focused on the genetic contributions to iron status. Population differences in the frequency of genetic variants that are associated with increased risk of iron disorders may explain varying iron status in different ethnic groups. No studies to date have pooled existing data to summarize current findings and knowledge on ethnic differences in iron status. This review summarizes data on genetic variants found to be associated with iron metabolism across different ethnic groups and statistically evaluates published data on iron status as a function of ethnicity. 112 Current Status of Knowledge Iron physiology Iron absorption Dietary iron is ingested as heme iron (from animal-based foods) and non-heme iron (from animal- and plant-based foods). Heme iron is only minimally impacted by iron stores. The proteins involved in heme iron absorption are unique to heme iron and the pathways involved in this process continue to be elucidated (16–18). In contrast, non-heme iron absorption is tightly regulated in response to body iron demands and mutations in the proteins involved in non-heme iron absorption are associated with known iron-related diseases (16,19,20). Dietary non-heme iron is ingested primarily as ferric iron (Fe3+) that must be reduced to ferrous iron (Fe2+) (by duodenal cytochrome b) before it enters the enterocyte via the divalent metal transporter 1 (DMT1). Once absorbed, Fe2+ can either be stored as ferritin, used for intracellular functions, or exported via the only known non-heme iron export protein in the body, ferroportin (FPN). Once exported, the ferroxidase hephaestin oxidizes Fe2+ to Fe3+, which can then bind to transferrin (TF) in the bloodstream (16,20,21). At other locations in the body, ceruloplasmin functions as a ferroxidase (22), and the placenta also expresses a unique ferroxidase (zyklopen) (23). Systemic iron homeostasis Transferrin-bound Fe3+ in circulation is taken up by cells using receptor mediated endocytosis via transferrin receptor 1 (TFR1). Cellular expression of TFR1 is highest in erythroid tissue to support erythropoiesis. Under normal circumstances, approximately 30% of the iron-binding sites in the plasma TF pool are occupied. This value is referred to as transferrin 113 saturation (TSAT) (16,24). When TSAT exceeds 45%, iron begins to circulate free or bound to low molecular weight molecules (citrate, albumin) generating potentially toxic iron species known as non-transferrin bound iron (NTBI). The pancreatic cells and hepatocytes can internalize NTBI via Zrt- and Irt-like protein 14, while cardiomyocytes are purported to internalize NTBI through L-type or T-type calcium channels (25,26). Cellular uptake of NTBI can increase the intracellular labile iron pool resulting in generation of reactive oxygen species that can cause oxidative damage, adversely impacting specific organs and over time, increasing the risk of chronic diseases. Only 1-2 mg/day of absorbed iron is needed to offset the typical amounts of endogenous daily iron losses. The majority of iron utilized to support erythropoiesis (20-25 mg/day) is obtained from catabolism of senescent RBCs. Several hormones are now known to be involved in the regulation of systemic iron homeostasis including hepcidin, erythropoietin, and erythroferrone. Hormonal control of iron physiology Of the three iron-related hormones, hepcidin is the major regulator of body iron balance. This hepatic hormone binds to FPN thereby reducing iron export from the enterocyte and iron release from stores (27). Hepcidin production is induced by iron loading or inflammation/infection and suppressed by increased erythropoietic demand or hypoxia. The regulation of hepcidin involves multiple membrane proteins (MT-2, matriptase-2; HJV, hemojuvelin; BMPR 1 and 2, bone morphogenetic protein receptor 1 and 2; TFR1 and 2; HFE, high iron protein; IL-6 receptor), intracellular signaling pathways (BMP-SMAD, BMP-small mothers against decapentaplegic; JAK-STAT, janus kinase-signal transducer and activator of 114 transcription protein), and hormones (erythropoietin, erythroferrone) (21,28–30). Genetic mutations in these proteins are associated with iron disorders as shown in Figure 4.1. Figure 4.1 The hepatic hepcidin signaling pathway and common genetic mutations in this signaling pathway. Hepcidin is regulated by infection/inflammation (left), iron status (middle), and erythropoietic demand (right). Pathogenic mutations in genes that encode proteins involved in the regulation of the hepcidin or FPN result in iron deficiency (green X) or iron overload (red X). When serum transferrin is saturated, diferric-transferrin binds to TFR1 and displaces HFE. HFE can then form a complex with TFR2 and possibly HJV to promote the BMP-SMAD hepcidin signaling pathway. Recessive mutations in the genes encoding these proteins (HFE, TFR2, HJV, hepcidin) result in decreased hepcidin production preventing hepcidin from being upregulated as iron stores accumulate leading to iron overload. Under low oxygen conditions, MT-2 cleaves HJV, generating a soluble and inactive form of this protein, resulting in the inactivation of the BMP-SMAD signaling pathway and downregulation of hepcidin transcription. Genetic mutations in MT-2 result in uninhibited hepcidin production leading to IRIDA. High 115 erythropoietic demand results in upregulation of erythroferrone, which suppresses hepcidin production by sequestering BMP6 and inhibiting BMP-SMAD signaling. In inflammation, IL-6 binds to IL-6 receptor, stimulating the JAK-STAT pathway and upregulating hepcidin production. Dominant gain- and loss-of-function FPN1 mutations result in hepcidin resistance or in cellular iron accumulation (particularly in macrophages), respectively. Recessive mutations in DMT1 result in iron loading anemia. Likewise, recessive mutations in TF result in reduced concentrations of functional TF and iron loading anemia. The figure was created using BioRender.com. BMP, bone morphogenetic protein; BMPR, bone morphogenetic protein receptor; DMT-1, divalent metal transporter 1; FPN, ferroportin; HAMP, hepcidin, HFE, high iron protein, HJV, hemojuvelin; IRIDA, iron refractory iron deficiency anemia; JAK-STAT, Janus kinase-signal transducer and activator of transcription protein; MT-2, matriptase-2; SMAD, small mothers against decapentaplegic; TF, transferrin; TFR, transferrin receptor. Assessment of iron status Multiple biomarkers are used to represent iron status of individuals and/or populations. Ferritin is the primary intracellular iron storage protein and some is released into circulation as serum ferritin (SF) (31). Low SF concentrations are indicative of depleted iron stores. Elevated concentrations of SF can indicate excessive iron stores or inflammatory conditions since ferritin is an acute phase protein (31). Total iron binding capacity (TIBC) indicates the total number of binding sites for iron in TF, which is the main iron transport protein in circulation. TIBC, in combination with serum iron, can be used to calculate TSAT, which reflects iron supply to tissues (31). Soluble transferrin receptor (sTfR) reflects the intensity of erythropoietic and cellular demands for iron (31). An increase in sTfR concentrations indicates an increase in cellular iron requirements or insufficient iron supply to tissues (31). In late stages of ID, hemoglobin production is compromised which can result in anemia (31). A summary of the typical cutoffs for ID and/or IO for common iron status indicators is provided in Appendix Table 1. 116 Genetic variants of iron metabolism Tight control of iron absorption is needed to prevent ID or IO but to date, only about 30% of the interindividual variability in iron absorption can be captured by hepcidin or other iron status biomarkers (32–36). Marked ethnic differences in the risk of ID or IO have been known to exist as highlighted by the CDC cut-offs used to denote anemia (37). To investigate the genetic contributions to variations in iron status between and within populations, multiple genome-wide association studies (GWAS) and candidate gene association studies have been undertaken. The following sections cover pathogenic mutations in iron-related genes identified to date that result in ID or IO, and polymorphisms within iron- and non-iron related genes found to be associated with variations in iron traits among healthy populations. Genetic variants associated with iron status in healthy populations stratified by ethnic group are summarized (Appendix Table 2). Genetic variants related to iron overload Genetic mutations in proteins involved in the regulation of the hepcidin/FPN axis result in aberrant expression of hepcidin or FPN (38) and can eventually lead to IO. Disorders of the hepcidin/FPN axis cause subtypes of hereditary hemochromatosis (HH), a heterogeneous genetic condition found primarily in Northern Europeans. The most common form of HH occurs from mutations in HFE, but HH can also result from mutations in HJV, hepcidin gene (HAMP), TFR2, and FPN1, which are collectively referred to as non-HFE hemochromatosis. The pattern of inheritance of pathogenic mutations in HFE, HJV, HAMP and TFR2 is autosomal recessive, whereas in FPN1 is autosomal dominant (39). While mutations in these iron-related genes have 117 been described as causative for HH, some single nucleotide polymorphisms (SNP) within, or in close-proximity to these genes, have been associated with iron traits in healthy populations. HFE HFE was named as the high iron (high Fe) or hemochromatosis protein when it was discovered in 1996 to harbor a mutation that was highly prevalent in patients with hemochromatosis (40). This major histocompatibility complex class I like protein stimulates the intracellular BMP-SMAD hepcidin signaling pathway. Recessive mutations in HFE result in decreased hepcidin production and thus, hepcidin is not appropriately up-regulated as iron stores accumulate (41) (Figure 4.1). HFE-related hemochromatosis or HH type 1 is most commonly associated with two missense mutations, C282Y and H63D. The iron burden and clinical phenotype presented in patients with HFE-related hemochromatosis is highly variable. The classic biochemical abnormalities seen in C282Y homozygous individuals include elevated TSAT and SF, and tissue IO. Although biochemical penetrance of homozygosity for this mutation is high (75-100% in males, 40-60% in females) (42–44), the clinical penetrance is much lower, affecting men at higher rates than women (2-38% and 1-10%, respectively) (45,46). Interestingly, H63D homozygosity rarely results in clinical disease development, except when this mutation is present with C282Y, both in heterozygote states (40,47). The prevalence of the C282Y and H63D mutations varies among ethnic groups and is one of the best examples of ethnic differences in iron metabolism (48). The C282Y homozygosity is most prevalent (0.3-0.5%) in individuals of Northern European descent (44,49–52). This mutation is thought to have originated in a Celtic population 60-70 generations ago (53–55). A Viking origin of this mutation has been proposed as highest frequencies are observed in populations of Northern European descent (i.e. Viking populations) (56). The Hemochromatosis 118 and Iron Overload Screening (HEIRS) study reported the highest prevalence (0.44%) of C282Y homozygosity among unrelated non-Hispanic white individuals (44). The prevalence of C282Y homozygosity has been reported at even higher frequencies in specific regions of Northern Europe, particularly Ireland (1.24-1.96%) (57,58). Additionally, the highest average allele frequency in HFE has been reported among this population (10.1%) (48). The lowest prevalence of C282Y homozygosity in the HEIRS study was reported among Asians (0.000039%), followed by Pacific Islanders (0.0132%), African Americans (0.014%), Hispanics (0.027%), and Native Americans (0.11%) (44). The H63D mutation has a broader distribution with higher frequencies throughout Europe and other geographical locations (48). The estimated prevalence of H63D homozygosity in the HEIRS study was 2.4% in non-Hispanic whites, 1.3 % in Native Americans, 1.1% in Hispanics, 0.089% in African Americans, and 0.02% in Pacific Islanders and Asians (44). The prevalence of C282Y/H63D compound heterozygosity reported in the HEIRS study was 2.0% in non-Hispanic whites, 0.77 % in Native Americans, 0.33% in Hispanics, 0.071% in African Americans, and 0.096% in Pacific Islanders, and 0.0055% in Asians (44). SNPs in HFE corresponding to C282Y and H63D have been associated with iron and erythrocyte traits in healthy populations. The SNP rs1800562 that results in the C282Y variant has been associated with several iron biomarkers (59–63) and hematological parameters (64) that reflect both systemic and cellular iron homeostasis at the genome-wide level among Europeans. The SNP rs1799945 that results in the H63D variant has also been associated at the genome- wide level with various iron traits (59,65) in Europeans. In Hispanics, both rs1800562 and rs1799945 have been shown to be associated with iron status biomarkers (66). In African Americans, only the association of rs1800562 with SF and TF seen in Europeans has been 119 replicated (67). Noteworthy, these variants are nearly absent in Asians and Pacific Islanders (44,52), and no associations between these mutations and iron biomarkers have been reported in these populations (68). HJV and HAMP The HJV encodes the hemojuvelin protein which acts as a BMP co-receptor to regulate the expression of HAMP (21) (Figure 4.1). Juvenile hemochromatosis (JH) or HH type 2 is the most severe form of HH (69,70) and can arise from pathogenic mutations in HJV (HH subtype 2a) or HAMP (HH subtype 2b) and results in cardiomyopathy, diabetes and hypogonadism by the early 20s (69,70). In rare instances adult-onset HH due to HJV mutations has been observed (71,72). Biochemical abnormalities of JH include high TSAT and marked increases in SF (70). Interestingly, a systematic review assessing the genotypic and phenotypic spectrum of HJV mutations in patients with HH reported ethnic disparities in the clinical presentation of HJV- related HH between Europeans and East Asians (73). This same review (73), in agreement with a phenotypic analysis of HH subtypes (70) found that European males and females were affected equally, while East Asian males were affected at higher rates than their female counterparts (73). The first comparative study of the prevalence of HFE and non-HFE related HH reported a predicted prevalence of homozygous HJV pathogenic mutations of 1 in 5 million using available next-generation sequence (NGS) databases, with the highest predicted prevalence in South Asians (74). Homozygous HAMP pathogenic mutations were predicted to be even rarer (1 in 182 million) (74). Of note, these predictions may be restricted given that the populations were not representative of all the pathogenic mutations within the HH-related genes identified to date. 120 Mutations in HJV account for up to 90% of JH cases (73), and most have been identified in a single family or small populations (73). The G320V mutation is a more common HJV mutation and is restricted to European ancestry (75,76). Other mutations have been described only within individuals of Asian or Pacific Island ancestry (73,77), some are more common among East Asians (e.g. C321X, Q312X) (71,78–85) and some among South Asians (e.g. G336X, G99R, P192L, L194, C80Y, A343PfsX23) (77,86,87). Few mutations have been described in Africans (e.g. R385X, A310G) (73,88,89), however, these are not restricted to African ancestry (90–93). Mutations in HAMP result in a rare form of JH and have been characterized in patients from varying geographical locations (87,93,102,94–101). Interestingly, no polymorphisms within the HJV or HAMP have been associated with variation in iron traits among healthy individuals. A candidate gene association study reported an association between the SNP rs10421768 in HAMP and hemoglobin among a Kenyan cohort (n = 628) (103), but the significance of this association disappeared after adjustment for multiple testing. TFR2 TFR2 forms a complex with HFE and possibly HJV to promote BMP-SMAD signaling and upregulate hepcidin production (29) (Figure 4.1). HH type 3 results from pathogenic mutations in TFR2 and is a disease mainly of adult onset, although, more severe mutations have been described in early childhood (104–106). Biochemical abnormalities include elevated TSAT and SF (24). Although the true prevalence of TFR2 mutations is unknown, homozygous TFR2 pathogenic genotypes have been estimated (using available NGS databases) to be 1 in 6 million 121 and were predicted to be most frequent among non-Finnish European populations from ExAC database (74). Most mutations in TFR2 that result in HH type 3 have been clustered in ~45 families worldwide (107) but some mutations have been described among unrelated individuals from different ethnic groups (105,108–111). TFR2-related HH may be the leading cause of IO in Asians, predominantly those from Japan (112,113). The I238M variant is present in Asians at a higher frequency (0.0192) (112), however, it is not restricted to Asian ancestry (93). The rs7385804 SNP in TFR2 has been associated iron biomarkers (59,64,65) and RBC parameters (59,114) at genome-wide significance in healthy Europeans. The associations of rs7385804 with serum iron and TSAT have been replicated in Hispanics (66) and Chinese women (115). Another SNP within TFR2, rs7786877, has been associated with mean corpuscular volume (MCV) at the genome-wide level among individuals of European ancestry (64). FPN The FPN1 gene encodes FPN, which regulates cellular export of non-heme iron (116) (Figure 4.1). Pathogenic mutations in FPN1 can result in two phenotypically distinct diseases, HH type 4A (FPN disease) and HH type 4B. HH type 4B is caused by gain-of-function mutations resulting in partial or complete hepcidin resistance (117). Individuals with these mutations present with high SF, TSAT, organ IO, and progressive organ damage (117). The phenotype of HH type 4B resembles other recessive HH-causing mutations (117). Conversely, FPN disease is due to dominant loss-of-function mutations that result in impaired iron export, particularly from reticuloendothelial cells (117). The classical phenotype of FPN disease includes high SF, normal to low TSAT, low hemoglobin concentrations, and progressive iron 122 loading (117,118). FPN disease seems to have a milder clinical presentation as it has not shown to cause major organ damage (119,120), conceivably because macrophages protect against reactive oxygen species when burdened with iron (117). Not all FPN mutations have been classified based on their phenotypic presentation or pathogenicity (117,119). Globally, the estimated pathogenic genotype carrier rate of FPN1 mutations (using available NGS databases) is 1 in 1,373 (74). The highest predicted pathogenic genotype carrier rate was reported in African Americans (0.25%), followed by Americans (0.039%), East Asians (0.033%), and non-Finnish Europeans (0.03%) (74). A systematic meta-analysis of FPN1 mutations found 31 disease-causing mutations in 161 individuals by 2010 (119), and more recently a total of 60 variants in 359 individuals was described between 1999 to 2019 (120). The most frequently reported FPN1 mutation is the V162del loss-of-function mutation (119–126). Few polymorphisms in FPN1 have been associated with iron traits in healthy populations. Q248H is the most common FPN1 variant among individuals of African ancestry and is present at polymorphic frequencies in African populations (2.2-13.4%) (127–131). The Q248H variant has been weakly associated with increased SF levels among individuals with primary IO and healthy individuals (127–129,131–133), and this association seems to be stronger in men (132,133). This polymorphism may confer a protective effect against ID (128,134), anemia, and iron deficiency anemia (IDA) (134) in African children, particularly those with underlying inflammation. While conflicting (134), a protective effect against malarial infection has also been suggested. The largest GWAS on iron status conducted in healthy Europeans identified a SNP (rs744653) near FPN1 that was associated with SF and TF at genome-wide significance (59), and the association with SF was replicated in a recent candidate gene association study among Europeans (135). Lastly, a GWAS conducted in healthy Chinese men 123 found a SNP (rs5742933) located in close proximity to the FPN1 gene that was associated with SF at the genome-wide level (136). However, this SNP seemed to be in weak linkage disequilibrium (R2 and D’<0.20) with SNPs located within FPN1 (136). Genetic variants related to iron deficiency As with IO, several SNPs in iron-related genes have been associated with increased risk of anemia, ID or IDA in healthy populations of various ethnicities. Hereditary disorders that result in anemia include iron-refractory iron deficiency anemia (IRIDA) and iron loading anemia. TMPRSS6 MT-2 is produced in the liver and negatively regulates expression of hepcidin by cleaving HJV and in turn, inactivating the BMP-SMAD pathway (137) (Figure 4.1). Genetic mutations in the MT-2 protein (TMPRSS6, transmembrane serine protease 6) can result in uninhibited hepcidin production leading to IRIDA (138–140). This disease is characterized by congenital hypochromic, microcytic anemia, low MCV and TSAT, low to normal SF levels, and defective iron absorption and utilization (138,141). IRIDA is usually unresponsive to oral iron, and only partially responsive to parenteral iron treatment (138). The prevalence of pathogenic mutations resulting in IRIDA is unknown (140–142). So far, 69 different TMPRSS6 mutations in 65 IRIDA families of different ethnicities have been identified (137,141–144). Most IRIDA patients have homozygous mutations in TMPRSS6 and thus, the mode of inheritance is considered to be recessive (143). However, heterozygous mutations may also result in a clinical phenotype that resembles IRIDA but with a milder 124 presentation (141,143). More recent evidence suggests that IRIDA is a highly heterogeneous disease and some patients have been found to respond to oral iron therapy (143). Several genome-wide and candidate gene association studies conducted in healthy populations of mainly non-African origin, have found common SNPs in TMPRSS6 to be associated with iron traits. The most reported SNP in TMPRSS6, rs855791 (V736A), has been associated with iron status indicators (59,62,145,146), RBC parameters (59,64,147), and liver iron content (148) in healthy individuals of European ancestry at genome-wide significance. Additionally, candidate gene studies in Europeans have replicated the association of this SNP with SF (135,149) and serum iron (149) and have identified associations with sTfR (61) and the sTfR:SF ratio (150). This SNP was also associated with hemoglobin (6,115) and iron status indicators (6,115) in East Asians, and with serum iron in Hispanics (66). A systematic review with meta-analysis found the minor allele (A) frequency (MAF) of rs855791 to be significantly higher in Asians than in Europeans (0.55 vs 0.42) (151). This same review showed that each A allele was associated with 0.11 g/dL lower hemoglobin concentrations, 3.71 µg/L lower SF levels, and 0.2 mg/L higher sTfR, and the differences in effect estimates between ethnicities were not significant (151). Significant differences in MAF between Africans and non-Africans have been reported (152). The rs855791 SNP has a significantly lower MAF in Africans (<0.1) compared to non-Africans (<0.35), yet Africans have a high prevalence of anemia (152). The second most reported SNP in TMPRSS6 is rs4820268 (D521D) and it has been associated with iron biomarkers (60,65,146), and hemoglobin in Europeans (147), with SF and hemoglobin in East Asians (6), and with hemoglobin in South Asians (147). The association with serum iron was replicated in a candidate gene study of Europeans and an association with SF was identified (149). A study in Europeans found the rs4820268 GG genotype to be associated with 125 lower serum iron, hemoglobin, MCV, and mean cell hemoglobin, and higher TF, sTfR and sTfR:SF ratio (65). Furthermore, a meta-analysis of 13 study populations showed that the G allele resulted in a 0.12 µg/L increase in SF concentrations among Europeans and a 3.69 µg/dL decrease among Asians (151). Analysis of 3 studies among Europeans suggested an association between the G allele and a reduction in sTfR (151). However, there was high heterogeneity of the effect of each allele on the variation in iron status parameters reported in studies among Europeans included in this meta-analysis (151). Consistent with these observations, a recent study assessing differences in allele frequencies among different populations found that South and East Asians had the highest number of iron lowering alleles and Africans had the lowest number of low iron risk alleles (152). Furthermore, studies in Chinese populations identified rs855791 and rs4820268 polymorphisms as genetic risk factors for developing anemia, ID, and IDA (6,115). Several other SNPs in TMPRSS6 have been associated with RBC parameters (64,147), and iron biomarkers (146) among different ethnic groups. TF Autosomal recessive mutations in TF cause severely reduced serum concentrations of functional TF and lead to hypotransferrinemia (or atransferrinemia) (Figure 4.1). This rare disorder is characterized by iron deficient erythropoiesis and anemia due to insufficient iron supply to erythropoietic tissues and severe IO in non-hematopoietic organs due to low hepcidin levels and increased non-TF mediated iron uptake (153,154). Only 18 cases among 16 families world-wide have been described to date (155,156). While mutations that result in hypotransferrinemia are extremely rare, several SNPs in TF have been associated with iron status in populations from different ethnic origin. The 126 rs3811647 SNP in TF has been associated with TIBC and TF in genome-wide and candidate gene association studies in Europeans (60,63,65,150,157). Other SNPs in TF have been associated with serum iron, TF, and TSAT (59). Associations of several polymorphisms in TF with TIBC found in Europeans have been replicated in Asians, Hispanics and/or African Americans (157). Additionally, a candidate gene association study in Chinese women found TF polymorphisms were significantly associated with serum iron, TF, and TSAT (115). Other SNPs in TF have been associated with TSAT and TIBC in Hispanics (66), and with TIBC in African Americans (67). Moreover, the GWAS in African Americans found the top two SNPs in TF to explain 11.2% of the variation in TIBC levels in this population (67). Interestingly, unlike other iron-related genes discussed in this section, in TF, no SNPs have been reported to be associated with hematological traits in any ethnic group. DMT1 DMT1 transports dietary iron into enterocytes (Figure 4.1) or out of intracellular endosomes. Pathogenic mutations in DMT1 result in microcytic anemia and severe organ iron loading (141). Despite having body iron excess, patients with defects in DMT1 have normal to mildly elevated SF levels (141). Iron loading anemia due to mutations in DMT1 is extremely rare with only 7 patients from 6 families with homozygous or compound heterozygous described to date (141). Despite the importance of DMT1 in iron trafficking, few polymorphisms have been associated with iron traits. In a healthy Turkish cohort, the IVS4+44 polymorphism in DMT1 was associated with inter-individual variations in serum iron (158). This same SNP was associated with up to a four-fold increased risk of developing anemia in Italian children with Celiac disease (159). 127 Other genetic variants related to iron metabolism Iron-related genes Polymorphisms in other iron-related genes involved in transport of non-heme and heme iron (TFR1; CYBRD1, cytochrome b reductase 1; FLVCR, feline leukemia virus subgroup C receptor-related protein; STEAP3, six-transmembrane epithelial antigen of prostate 3; CD163, cluster of differentiation 163), in the regulation of cellular hepcidin signaling pathways (SMAD8; BMP2; BMP4; BMP9; BMPR1B; BMPR2; NEO1, neogenin 1; PCSK7, protein convertase suntilisin/kexin type 7), or intracellular iron signaling (HIF2A, hypoxia-inducible factor 2 alpha; IRP1, iron regulatory protein 1), have been associated with or have suggestive associations with at least one biomarkers of iron status and/or erythrocyte phenotype in healthy populations (59,61,135) or act as modifiers of HH phenotype (160–162) among individuals of European descent. Non-iron related genes Although most genetic variants associated with iron status indicators are within or in close proximity to iron-related genes, genetic variants of non-iron related genes have also been identified. Among Europeans, SNPs in genes involved in lipid metabolism were shown to be associated with TF (NAT2, N-acetyltransferase 2; ARNTL, aryl hydrocarbon receptor nuclear translocator like; FADS2, fatty acid desaturase 2), SF (ABO, alpha 1-3-N- acetylgalactosaminyltransferase and alpha 1-3-galactosyltransferase; TEX14, testis expressed 14) (59) or serum iron (LPL, lipoprotein lipase) (146). In Hispanics, a SNP in myelin regulatory factor (MYRF) that is in linkage disequilibrium with the FADS2 SNP identified in Europeans 128 (59) was associated with TIBC (66). Moreover, the associations between SNPs in NAT2 with TF and in ABO with SF found in Europeans (59) were generalized to Hispanics (66). Whether such associations result from the confounding influence of variations in lipid metabolism on iron status or reveal a pleiotropic effect of lipid-related genes on iron regulation needs further investigation. Interestingly, in analysis of the UK Biobank samples, ferritin-associated SNPs in HFE and TMPRSS6 conferred significant protection against hypercholesterolemia, suggesting interplay of metabolic pathways between lipid and iron (163). Other SNPs in non-iron related genes have been associated with iron traits, including TIBC, unsaturated iron binding capacity, serum iron, and sTfR in Europeans (146,157). In African Americans, SNPs in the hepatoma-derived growth factor-like protein 1 (HDGFL1) and MAF bZIP transcription factor (MAF) have been associated with TIBC and in growth factor receptor bound protein 2-associated protein 3 (GAB3) with SF at the genome-wide significance (67). Among these associations, only the significant SNP in GAB3 was replicated in Hispanics (66). GAB3 is a member of the GBR2-associated binding protein gene family and is involved in several growth factor and cytokine signaling pathways. The protein encoded by this gene is expressed in hematopoietic tissues and facilitates macrophage differentiation (164,165). In East Asians, a SNP in postmeiotic segregation increased 1 (PMS1) was found to be significantly associated with SF (136). PMS1 encodes a protein thought to be involved in DNA mismatch repair pathways and is expressed by various tissues, including hematopoietic tissues. Interestingly, anti-PMS1 antibodies were found in Japanese patients with aplastic anemia (10%) but none were detected in aplastic anemia patients from the Unites States (166). Additionally, while PMS1 has not been shown to play a direct role in iron homeostasis, it is in close proximity to FPN1. 129 Ethnic differences in iron status Iron regulation is highly conserved, but risk of iron related disorders differs across major ethnic groups. These ethnic differences may be a consequence of adaptive changes during evolution that occurred due to limited iron availability, a condition referred to as antagonistic pleiotropy. In antagonistic pleiotropy, adaptive changes that arise from evolutionary adaptations can become deleterious in the current environment. For iron, due to its importance in physiology, strong selective pressures would have been expected to maintain adequate iron status. These adaptations needed in an iron poor environment, i.e. mutations that increase iron absorption, may now be deleterious when iron is considerably more abundant in our food systems. Understanding the evolutionary context underlying dietary adaptation could have strong implications in precision nutrition (167). Large epidemiological reports of ethnic differences in iron status indicators have been published over the past few decades. Several studies of North American cohorts have evaluated iron status in large groups of otherwise healthy adults. The largest epidemiological study to date that evaluated iron stores as a function of ethnicity was the HEIRS study. This study recruited 101,168 primary care adults aged 25 years or older from the United States and Canada, and evaluated SF and TSAT as a function of ethnicity (self-reported as Hispanic, European, African American, Asian, Pacific Islander, Native American) (168). Another large epidemiological study, the Recipient Epidemiology and Donor Evaluation Study-III recruited 12,683 men and women participants of different ethnicities (self-reported as African American, Asian, White, Hispanic, and other) aged 18 years or older that had successfully donated whole blood (169). This study was designed to examine the genetic and metabolic basis of blood donor susceptibility to ID and iron-related symptoms in multiple ethnic groups. The Iron and Atherosclerosis Study 130 (FeAST) was a prospective, randomized controlled single-blinded clinical trial to test whether iron reduction using phlebotomy in participants with symptomatic but stable peripheral arterial disease can effectively improve clinical outcomes. This study recruited 1,277 European and African American veterans over the age of 21 (primarily males) (170). An additional study among 1,491 African American and 31,005 European men and women compared hematologic and iron status between these two groups (171). Another study designed to determine the frequency of HFE mutations and its association with iron related genotypes involved 10,198 adults (self-reported as Black, Asian, White, and Hispanic) (172). Additional epidemiological data examining iron status among multiple ethnic groups was conducted using the NHANES III database involving 20,040 individuals aged 18 years or older (grouped as Black, White, Hispanic, and other) (173). Similarly, large cohort data on iron status have been published in Europe. The population-based SUNSET study was a multiethnic, cross-sectional study designed to test the association between SF and the prevalence of type 2 diabetes and fasting glucose concentrations in a total of 2,975 adults (174). In addition, two other cohort studies targeting women of reproductive age were conducted to examine ethnic differences in iron status (175,176). Major findings from these data are summarized below by ethnicity. Europeans Most research to date has been conducted in Europeans, which are often used as the reference group in comparisons to other ethnic groups. Europeans have been consistently shown to have a lower risk of ID and anemia, and higher hemoglobin concentrations (171,177–180). In addition, TSAT in Europeans has been found to be significantly higher than mean values observed in African Americans and Hispanics (179,180). 131 African Americans Lower hemoglobin concentrations are consistently reported among African Americans compared to Europeans. These differences have been noted in infancy and appear to be maintained in the elderly (171,176,181–184). Existing data from population groups with sample sizes ranging from 388 to 3,074 indicate that the hemoglobin distribution observed among African Americans is shifted to the left by approximately 0.8 g/dL (184,185). This difference has been highlighted by the CDC and WHO to promote ethnicity-specific cutoffs for hemoglobin in the diagnosis of anemia (37,186). Ethnic differences in the hemoglobin distribution remain significant even after controlling for iron status (SF and TSAT) (182,184) and/or dietary iron intake (182), suggesting that variation in hemoglobin concentration is not entirely driven by factors related to iron metabolism. Despite lower hemoglobin concentrations, African American adults have elevated SF concentrations compared to Europeans and Hispanics (44,170,173,183,186–188). NHANES III (173,189), FeAST (170,188), and HEIRS (179) data all found significantly higher SF in African Americans compared to Europeans. However, both NHANES and HEIRS (179,180), and Li et al.’s genetic study in African Americans (67) found that African ancestry was associated with decreased levels of serum iron and TSAT as well as increased levels of sTfR. Moreover, prevalence of ID and anemia in HEIRS was greater in African Americans than in Europeans and Asians among women of reproductive age (177,190,191). At present, whether the elevated SF concentrations in African Americans reflects increased systemic inflammation or elevated iron stores is unclear. African American populations are often at increased risk for obesity compared to other ethnic groups (192), and elevated BMI is positively correlated with SF concentrations, perhaps due to adiposity-induced inflammation. The lower TSAT and 132 hemoglobin concentrations but higher SF levels observed in African Americans might indicate that the mobilization of iron from stores for erythropoiesis is reduced due to some unknown genetic contributors (182). East Asians Asian populations can be further sub-divided into East Asians or South Asians based on geographical distribution and ethnicity. Although data in Asian groups are not as abundant, numerous studies have reported that East Asian populations exhibited higher iron stores and a higher risk of IO. This finding is evident despite the fact that the frequencies of the most common HFE mutations are lowest among Asians (44). Of note, HEIRS individuals that self- identified as Asians (predominantly East Asians: Chinese, Japanese, Vietnamese, and Filipino) exhibited the highest SF, TSAT or both SF and TSAT compared to any other population group studied, even after excluding polymorphisms associated with IO (44,191,193). These differences remained significant after adjusting for diabetes or liver disease (193). In a further analysis of HEIRS data focused on women of reproductive age, Asian ancestry was associated with a decreased risk of ID and with increased iron stores compared to other ethnic groups independent of known HH mutation in HFE (C282Y) (191). Additional epidemiologic data from a study in Korean adults (n = 4,904) found mean TSAT in both females and males was significantly higher than in Europeans as reported in the HEIRS study (194,195). Few data exist on possible mechanisms explaining these ethnic differences. A recent functional study evaluated iron absorption in a group of young East Asian women (32) and found mean percent iron absorption was significantly higher in East Asian women compared to that reported in European women using the same methodology, even after correcting for a fixed amount of SF (32). This observation suggests that the storage threshold at which East Asians down-regulate iron 133 absorption is higher, supportive of increased risk of IO in East Asian populations at maturity. This may also explain the observation that East Asian populations have a greater risk of diabetes at a lower BMI (196,197). South Asians Fewer data are available from South Asian populations, but existing data in this group suggests a lower iron status compared to Europeans and East Asians. Data from South Asian Surinamese adults (Hindustani Surinamese, n = 399) aged 35-60 years reported a significantly lower SF concentrations than observed among a cohort of Dutch adults (n = 508) (174). South Asian pregnant women (primarily from Pakistan and Sri Lanka, n = 198) were reported to have the highest risk of ID and anemia when compared to pregnant European (n = 326) and East Asian women (n = 43) (177). Lower hemoglobin, SF, and TSAT have also been noted in a group of South Asian women of reproductive age (176). Hispanics At least two epidemiological studies have reported a significantly higher prevalence of ID in Hispanics (178,190) compared to other ethnic groups, and Hispanic ethnicity has been associated with an increased risk of ID (191). In addition, evidence from NHANES found that Hispanic women of reproductive age had significantly lower SF and TSAT compared to the European women (180). Native Americans Data on iron status among Native American populations are limited. The HEIRS study included data from 645 Native Americans. In this relatively small cohort, mean TSAT and SF 134 concentrations in Native American men and women did not appreciably differ from the respective mean values reported in European men and women (198). Statistical evaluation of population-based data on iron status Multiple published reports of ethnicity and iron status exist but few attempts have been made to compile these data to explore statistical patterns of altered iron status. To address this gap, relevant population-based data were identified through PubMed, Web of Science, and Scopus using the following key words: ethnicity, race, genetic, Asian, European, Caucasian, African American, Chinese, Korean, genetics, iron, iron status, iron homeostasis, iron absorption, iron metabolism. Additional studies were identified through references cited within relevant articles. To systematically compare the data summarized above on iron status among different ethnical groups, we extracted iron status data from both population-based studies involving multiple ethnicities and other epidemiological or observational studies reporting data from one ethnic group only. We categorized studies by the number of ethnic groups involved and by the study type (Appendix Table 3). For analytical purposes, we excluded studies designed to assess iron status in frequent blood donors or pregnant and/or breastfeeding women, as this would be expected to impact iron status. Studies were also excluded if there were no available iron status data to extract or if the iron status data were not reported by sex. As for different literature examining the same study cohort, data from the largest sample size were included in the analyses. To explore possible differences in iron status indicators (SF, TSAT and hemoglobin) between different ethnic groups, linear mixed-effects models were constructed with ethnicity (East Asian, European, African American and South Asian) as a fixed-effect variable. Studies 135 where the data were collected from were considered as a random-effect variable. Sample size of each ethnic group in these studies was used as weight for the fixed-effect variable and the mean age of each ethnic group was controlled in the analysis. Mean age could not be controlled for models examining hemoglobin as this restricted the size of the dataset because many of the available studies did not report the mean age. Estimated marginal means of each iron status indicator were calculated for all ethnic groups using the package emmeans in R. All statistical analyses were performed using R version 3.4.3. The estimated marginal means of iron status indicators evaluated are presented in the Table 4.1. A statistical evaluation of mean differences in each indicator between ethnic groups is presented in Table 4.2. With this approach, Asians (predominantly East Asians) were found to exhibit significantly higher SF concentrations and TSAT when compared to Europeans, African Americans or South Asians, and this difference was significant for both males and females. While significant effects were evident in both sexes, the magnitude of the observed difference was significantly higher in males. This finding is expected as women of reproductive age have monthly losses of iron from menses, which may partially explain why risk of excess iron accumulation is greater in males. Evidence has shown that 1 µg/L of SF is equivalent to 8 mg of iron stores (199), thus the mean difference of 180 µg/L in SF would be translated into an additional 1440 mg of storage iron. If 1.5 mg of iron is absorbed daily (21), this additional 1440 mg of storage iron would be equivalent to the net amount of iron typically absorbed over 2.6 years. Hemoglobin concentrations are impacted only when iron stores have been depleted (200). African American men were found to have significantly lower hemoglobin compared to European, Asian or Hispanic men. Similarly, African American women had significantly lower 136 hemoglobin when compared to European women. These findings are consistent with previous literature evaluating hemoglobin as a function of ethnicity (171,176,181–185). Table 4.1 Mean reported iron status indicators in different ethnic groups1 Men Women SF, µg/L TSAT, % Hb, g/dL SF, µg/L TSAT, % Hb, g/dL 249 ± 17 37 ± 1 15.1 ± 0.1 113 ± 7 29 ± 1 13.1 ± 0.1 Asian [14] [6] [8] [13] [7] [9] (40,874) (22,543) (20,265) (31,036) (13,075) (9,706) 68 ± 16 32 ± 1 14.8 ± 0.1 27 ± 7 25 ± 1 13.5 ± 0.1 European [15] [8] [10] [20] [10] [13] (48,115) (38,532) (28,169) (62,736) (45,433) (33,352) 87 ± 16 30 ± 1 14.2 ± 0.1 45 ± 7 22 ± 1 12.7 ± 0.2 African [9] [8] [5] [8] [7] [4] American (14,025) (13,839) (2,882) (20,413) (19,908) (1,350) 57 ± 35 32 ± 5 35 ± 16 18 ± 2 12.6 ± 0.5 South Asian [1] [1] - [2] [2] [1] (150) (150) (488) (242) (188) 151 ± 83 29 ± 4 15.0 ± 0.2 54 ± 35 25 ± 4 13.3 ± 0.3 Hispanic [2] [1] [1] [2] [1] [1] (5,549) (5,122) (427) (7,784) (7,241) (543) 1 Data are presented as estimated marginal means ± SE [number of studies included in the analyses] (number of people included). The estimated marginal means were calculated from linear mixed-effects models using the lmer and emmeans functions. Hb, hemoglobin; SF, serum ferritin; TSAT, transferrin saturation. 137 Table 4.2 Comparison of iron status indicators between comparison ethnic groups and reference ethnic groups1 Reference Comparison Men Women Iron ethnic ethnic Differences Differences indicator N2 N2 group groups ± SE ± SE Asian [14] (40,874) [13] (31,036) European -180 ± 53 [15] (48,115) -86 ± 33 [20] (62,736) SF, African µg/L -161 ± 73 [9] (14,025) -68± 33 [8] (20,413) American South Asian -192 ± 303 [1] (150) -78 ± 163 [2] (488) European [15] (48,115) [20] (62,736) African 19 ± 44 [9] (14,025) 19 ± 23 [8] (20,413) American Asian [6] (22,543) [7] (13,075) European -5 ± 13 [8] (38,532) -3 ± 03 [10] (45,433) African -7 ± 13 [8] (13,839) -6 ± 03 [7] (19,908) TSAT, % American South Asian -5 ± 5 [1] (150) -11 ± 23 [2] (242) European [8] (38,532) [10] (45,433) South Asian -1 ± 5 [1] (150) -7 ± 25 [2] (242) African -2 ± 16 [8] (13,839) -3 ± 03 [7] (19,908) American African [5] (2,882) [4] (1350) American Hb, g/dL Asian 0.8 ± 0.13 [8] (20,265) 0.4 ± 0.2 [9] (9,760) European 0.6 ± 0.13 [10] (28,169) 0.8 ± 0.23 [13] (33,352) Hispanic 0.8 ± 0.13 [1] (427) 0.6 ± 0.3 [1] (543) Asian [8] (20,265) [9] (9,760) European -0.2 ± 0.1 [10] (28,169) 0.4 ± 0.17 [13] (33,352) 1The estimated differences were calculated from linear mixed-effects models using the lmer and emmeans functions. P-values reported were correction for multiple comparisons with Tukey's test. Hb, hemoglobin; SF, serum ferritin; TSAT, transferrin saturation. 2N, [number of studies included in the analyses] (number of people included) 3Significant difference between the comparison group and the reference group, P < 0.0001. 4Significant difference between the comparison group and the reference group, P = 0.0002. 5Significant difference between the comparison group and the reference group, P = 0.005. 6Difference between the comparison group and the reference group approached significance, P = 0.05. 7Significant difference between the comparison group and the reference group, P = 0.03. 138 Conclusions This review highlighted the shared and unique genetic variants among ethnic groups that have been associated with iron status biomarkers, and possible differences in iron status as a function of ethnicity were explored using published data. To date, most genome-wide and candidate gene association studies on iron homeostasis have been conducted in European populations. Some of the associations found in Europeans have been replicated in other ethnic groups but the clinical significance is unknown given the varying minor or effect allele frequencies among different populations. Interestingly, reported frequencies of key genetic variants associated with iron traits among different ethnic groups do not fully reflect epidemiological data on iron status in different populations. Because iron traits may be influenced by a combination of genetic, dietary, and lifestyle factors, methods taking into account several variants together, such as polygenic risk scores, may be better at predicting the risk of ID or IO in specific populations. Moreover, although recent studies have identified variants in non-iron related genes that are associated with iron traits, how they vary as a function of ethnicity and the significance of these SNPs has yet to be determined. Finally, most genome- wide and candidate gene association studies have been focused on blood biomarkers of iron status. Although iron status biomarkers correlate with body iron stores, their levels are sensitive to diet and disease, which may have confounded current association studies. Iron homeostasis in humans depends predominantly on the tight regulation of dietary iron absorption by enterocytes. Therefore, attempts to characterize genetic determinants of interindividual variation in iron homeostasis using direct measures of nutrient utilization are needed in order to better understand the differences in iron homeostasis that exist among different populations. Additional data are needed to identify possible risks and benefits associated with universal iron supplementation 139 policies such as those currently recommended for pregnant North American women (201), to identify the genetic basis of population differences in iron metabolism and disease susceptibility, and to help inform population-specific dietary iron intake recommendations and surveillance in at-risk populations. 140 REFERENCES 1. Lynch S, Pfeiffer CM, Georgieff MK, Brittenham G, Fairweather-Tait S, Hurrell RF, McArdle HJ, Raiten DJ. Biomarkers of Nutrition for Development (BOND)—Iron Review. J Nutr 2018;148(suppl_1):1001S-1067S. 2. Dev S, Babitt JL. Overview of iron metabolism in health and disease. Hemodial Int 2017;21 Suppl 1(Suppl 1):S6–S20. 3. Frey PA, Reed GH. The Ubiquity of Iron. ACS Chem Biol 2012;7(9):1477–1481. 4. Camaschella C. Iron deficiency. Blood 2019;133(1):30–39. 5. Montonen J, Boeing H, Steffen A, Lehmann R, Fritsche A, Joost H-G, Schulze MB, Pischon T. Body iron stores and risk of type 2 diabetes: results from the European Prospective Investigation into Cancer and Nutrition (EPIC)-Potsdam study. Diabetologia 2012;55(10):2613–2621. 6. Gan W, Guan Y, Wu Q, An P, Zhu J, Lu L, Jing L, Yu Y, Ruan S, Xie D, et al. Association of TMPRSS6 polymorphisms with ferritin, hemoglobin, and type 2 diabetes risk in a Chinese Han population. Am J Clin Nutr 2012;95(3):626–632. 7. Gujja P, Rosing DR, Tripodi DJ, Shizukuda Y. Iron overload cardiomyopathy: better understanding of an increasing disorder. J Am Coll Cardiol 2010;56(13):1001–12. 8. Eftekhari MH, Mozaffari-Khosravi H, Shidfar F, Zamani A. Relation between Body Iron Status and Cardiovascular Risk Factors in Patients with Cardiovascular Disease. Int J Prev Med 2013;4(8):911–6. 9. Basuli D, Stevens RG, Torti FM, Torti S V. Epidemiological associations between iron and cardiovascular disease and diabetes. Front Pharmacol 2014;5:117. 10. Torti S V., Torti FM. Iron and cancer: more ore to be mined. Nat Rev Cancer 141 2013;13(5):342–355. 11. Thévenod F. 15. Iron and Its Role in Cancer Defense: A Double-Edged Sword. Met Dev Action Anticancer Agents 2018;437–468. 12. Daugherty AM, Raz N. Appraising the Role of Iron in Brain Aging and Cognition: Promises and Limitations of MRI Methods. Neuropsychol Rev 2015;25(3):272–87. 13. Ward RJ, Dexter DT, Crichton RR. Neurodegenerative diseases and therapeutic strategies using iron chelators. J Trace Elem Med Biol 2015;31:267–273. 14. Simcox JA, McClain DA. Iron and diabetes risk. Cell Metab 2013;17(3):329–41. 15. Aregbesola A, Voutilainen S, Virtanen JK, Aregbesola A, Tuomainen T-P. Serum hepcidin concentrations and type 2 diabetes. World J Diabetes 2015;6(7):978–82. 16. Anderson GJ, Frazer DM. Current understanding of iron homeostasis. Am J Clin Nutr 2017;106(Supplement 6):1559S-1566S. 17. Gräsbeck R, Kouvonen I, Lundberg M, Tenhunen R. An Intestinal Receptor for Heme. Scand J Haematol 1979;23(1):5–9. 18. Tenhunen R, Gräsbeck R, Kouvonen I, Lundberg M. An intestinal receptor for heme: Its partial characterization. Int J Biochem 1980;12(5–6):713–716. 19. Knutson MD. Iron transport proteins: Gateways of cellular and systemic iron homeostasis. J Biol Chem 2017;292(31):12735–12743. 20. NISHITO Y, KAMBE T. Absorption Mechanisms of Iron, Copper, and Zinc: An Overview. J Nutr Sci Vitaminol (Tokyo) 2018;64(1):1–7. 21. Camaschella C, Nai A, Silvestri L. Iron metabolism and iron disorders revisited in the hepcidin era. Haematologica 2020;105(2):260–272. 22. Vashchenko G, MacGillivray RTA. Multi-copper oxidases and human iron metabolism. 142 Nutrients 2013;5(7): 2289–2313. 23. Chen H, Attieh ZK, Syed BA, Kuo YM, Stevens V, Fuqua BK, Andersen HS, Naylor CE, Evans RW, Gambling L, et al. Identification of zyklopen, a new member of the vertebrate multicopper ferroxidase family, and characterization in rodents and human cells. J Nutr 2010;140(10): 1728–1735. 24. Fleming RE, Ponka P. Iron Overload in Human Disease. N Engl J Med 2012;366(4):348– 359. 25. Jenkitkasemwong S, Wang C-Y, Coffey R, Zhang W, Chan A, Biel T, Kim J-S, Hojyo S, Fukada T, Knutson MD. SLC39A14 Is Required for the Development of Hepatocellular Iron Overload in Murine Models of Hereditary Hemochromatosis. Cell Metab 2015;22(1):138–150. 26. Chen M, Cabantchik ZI, Chan S, Chan GC, Cheung Y. Iron overload and apoptosis of HL-1 cardiomyocytes: effects of calcium channel blockade. PLoS One 2014;9(11):e112915. 27. Aschemeyer S, Qiao B, Stefanova D, Valore E V, Sek AC, Ruwe TA, Vieth KR, Jung G, Casu C, Rivella S, et al. Structure-function analysis of ferroportin defines the binding site and an alternative mechanism of action of hepcidin. Blood 2018;131(8):899–910. 28. Sangkhae V, Nemeth E. Regulation of the iron homeostatic hormone hepcidin. Adv Nutr 2017;8(1):126–136. 29. Muckenthaler MU, Rivella S, Hentze MW, Galy B. A Red Carpet for Iron Metabolism. Cell 2017;168(3):344-361. 30. Ganz T. Erythropoietic regulators of iron metabolism. Free Radic Biol Med 2019;133:69– 74. 143 31. World Health Organization. Assessing the iron status of populations. Geneva, Switzerland: World Health Organization, 2012. 32. Ye K, Cao C, Lin X, O’Brien KO, Gu Z. Natural selection on HFE in Asian populations contributes to enhanced non-heme iron absorption. BMC Genet 2015;16(1):61. 33. Young MF, Glahn RP, Ariza-Nieto M, Inglis J, Olbina G, Westerman M, O’Brien KO. Serum hepcidin is significantly associated with iron absorption from food and supplemental sources in healthy young women. Am J Clin Nutr 2009;89(2):533–538. 34. Roe MA, Heath ALM, Oyston SL, Macrow C, Hoogewerff JA, Foxall R, Dainty JR, Majsak-Newman G, Willis G, Fairweather-Tait SJ. Iron absorption in male C282Y heterozygotes. Am J Clin Nutr 2005;81(4):814–821. 35. Zimmermann MB, Troesch B, Biebinger R, Egli I, Zeder C, Hurrell RF. Plasma hepcidin is a modest predictor of dietary iron bioavailability in humans, whereas oral iron loading, measured by stable-isotope appearance curves, increases plasma hepcidin. Am J Clin Nutr 2009;90(5):1280–1287. 36. Roe MA, Collings R, Dainty JR, Swinkels DW, Fairweather-Tait SJ. Plasma hepcidin concentrations significantly predict interindividual variation in iron absorption in healthy men. Am J Clin Nutr 2009;89(4):1088–1091. 37. CDC. Recommendations to Prevent and Control Iron Deficiency in the United States. Clin Nurse Spec 1998;12(5):210. 38. Arezes J, Nemeth E. Hepcidin and iron disorders: new biology and clinical approaches. Int J Lab Hematol 2015;37(S1):92–98. 39. Brissot P, Pietrangelo A, Adams PC, Graaff B De, McLaren CE, Loreál O. Haemochromatosis. Nat Rev Dis Prim 2018;4:18016. 144 40. Feder JN, Gnirke A, Thomas W, Tsuchihashi Z, Ruddy DA, Basava A, Dormishian F, Domingo R, Ellis MC, Fullan A, et al. A novel MHC class I-like gene is mutated in patients with hereditary haemochromatosis. Nat Genet 1996;13(4):399–408. 41. Wu XG, Wang Y, Wu Q, Cheng WH, Liu W, Zhao Y, Mayeur C, Schmidt PJ, Yu PB, Wang F, et al. HFE interacts with the BMP type I receptor ALK3 to regulate hepcidin expression. Blood 2014;124(8): 1335–1343. 42. Beutler E, Felitti VJ, Koziol JA, Ho NJ, Gelbart T. Penetrance of 845G → A (C282Y) HFE hereditary haemochromatosis mutation in the USA. Lancet 2002;359(9302):211– 218. 43. Olynyk JK, Cullen DJ, Aquilia S, Rossi E, Summerville L, Powell LW. A population- based study of the clinical expression of the hemochromatosis gene. N Engl J Med 1999;341(10):718–724. 44. Adams PC, Reboussin DM, Barton JC, McLaren CE, Eckfeldt JH, McLaren GD, Dawkins FW, Acton RT, Harris EL, Gordeuk VR, et al. Hemochromatosis and iron-overload screening in a racially diverse population. N Engl J Med 2005;352(17):1769-78. 45. Grosse SD, Gurrin LC, Bertalli NA, Allen KJ. Clinical penetrance in hereditary hemochromatosis: Estimates of the cumulative incidence of severe liver disease among HFE C282Y homozygotes. Genet Med 2018;20(4):383–389. 46. Allen KJ, Gurrin LC, Constantine CC, Osborne NJ, Delatycki MB, Nicoll AJ, McLaren CE, Bahlo M, Nisselle AE, Vulpe CD, et al. Iron-Overload–Related Disease in HFE Hereditary Hemochromatosis. N Engl J Med 2008;358(3):221–230. 47. Camaschella C. Understanding iron homeostasis through genetic analysis of hemochromatosis and related disorders. Blood 2005;106(12):3710–3717. 145 48. Merryweather-Clarke AT, Pointon JJ, Jouanolle AM, Rochette J, Robson KJH. Geography of HFE C282Y and H63D mutations. Genet Test 2000;4(2):183–198. 49. Pietrangelo A. Hereditary hemochromatosis - A new look at an old disease. N Engl J Med 2004;350(23):2383-97. 50. Steinberg KK, Cogswell ME, Chang JC, Caudill SP, McQuillan GM, Bowman BA, Grummer-Strawn LM, Sampson EJ, Khoury MJ, Gallagher ML. Prevalence of C282Y and H63D mutations in the hemochromatosis (HFE) gene in the United States. J Am Med Assoc 2001;285(17):2216–2222. 51. Distante S, Berg JP, Lande K, Haug E, Bell H. High prevalence of the hemochromatosis- associated Cys282Tyr HFE gene mutation in a healthy Norwegian population in the City of Oslo, and its phenotypic expression. Scand J Gastroenterol 1999;34(5):529–534. 52. Hanson EH, Imperatore G, Burke W. HFE gene and hereditary hemochromatosis: A HuGE review. Am J Epidemiol 2001;154(3):193–206. 53. Simon M, Alexandre JL, Fauchet R, Genetet B, Bourel M. The genetics of hemochromatosis. Prog Med Genet 1980;4:135–168. 54. Smith BN, Kantrowitz W, Grace ND, Greenberg MS, Patton TJ, Ookubo R, Sorger K, Semeraro JG, Dovle JR, Cooper AG, et al. Prevalence of hereditary hemochromatosis in a Massachusetts Corporation: Is Celtic origin a risk factor? Hepatology 1997;25(6):1439– 1446. 55. Lucotte G. Celtic origin of the C282Y mutation of hemochromatosis. Blood Cells, Mol Dis 1998;24(4):433–438. 56. Milman N, Pedersen P. Evidence that the Cys282Tyr mutation of the HFE gene originated from a population in Southern Scandinavia and spread with the Vikings. Clin Genet 146 2003;64(1):36-47 57. Murphy S, Curran MD, McDougall N, Callender ME, O’Brien CJ, Middleton D. High incidence of the Cys 282 Tyr mutation in the HFE gene in the Irish population - implications for haemochromatosis. Tissue Antigens 1998;52(5):484–488. 58. Ryan E, O’Keane C, Crowe J. Hemochromatosis in Ireland and HFE. Blood Cells, Mol Dis 1998;24(4):428–432. 59. Benyamin B, Esko T, Ried JS, Radhakrishnan A, Vermeulen SH, Traglia M, Gögele M, Anderson D, Broer L, Podmore C, et al. Novel loci affecting iron homeostasis and their effects in individuals at risk for hemochromatosis. Nat Commun 2014;5(1):4926. 60. Benyamin B, McRae AF, Zhu G, Gordon S, Henders AK, Palotie A, Peltonen L, Martin NG, Montgomery GW, Whitfield JB, et al. Variants in TF and HFE Explain ∼40% of Genetic Variation in Serum-Transferrin Levels. Am J Hum Genet 2009;84(1): 60–65. 61. Oexle K, Ried JS, Hicks AA, Tanaka T, Hayward C, Bruegel M, Gögele M, Lichtner P, Müller-Myhsok B, Döring A, et al. Novel association to the proprotein convertase PCSK7 gene locus revealed by analysing soluble transferrin receptor (sTfR) levels. Hum Mol Genet 2011;20(5):1042–1047. 62. Traglia M, Girelli D, Biino G, Campostrini N, Corbella M, Sala C, Masciullo C, Viganò F, Buetti I, Pistis G, et al. Association of HFE and TMPRSS6 genetic variants with iron and erythrocyte parameters is only in part dependent on serum hepcidin concentrations. J Med Genet 2011;48(9):629–634. 63. McLaren CE, Garner CP, Constantine CC, McLachlan S, Vulpe CD, Snively BM, Gordeuk VR, Nickerson DA, Cook JD, Leiendecker-Foster C, et al. Genome-wide association study identifies genetic loci associated with iron deficiency. PLoS One 147 2011;6(3):e17390. 64. Ganesh SK, Zakai NA, Van Rooij FJA, Soranzo N, Smith A V., Nalls MA, Chen MH, Kottgen A, Glazer NL, Dehghan A, et al. Multiple loci influence erythrocyte phenotypes in the CHARGE Consortium. Nat Genet 2009;41(11):1191–1198. 65. Pichler I, Minelli C, Sanna S, Tanaka T, Schwienbacher C, Naitza S, Porcu E, Pattaro C, Busonero F, Zanon A, et al. Identification of a common variant in the TFR2 gene implicated in the physiological regulation of serum iron levels. Hum Mol Genet 2011;20(6):1232–1240. 66. Raffield LM, Louie T, Sofer T, Jain D, Ipp E, Taylor KD, Papanicolaou GJ, Avilés-Santa L, Lange LA, Laurie CC, et al. Genome-wide association study of iron traits and relation to diabetes in the Hispanic Community Health Study/Study of Latinos (HCHS/SOL): Potential genomic intersection of iron and glucose regulation? Hum Mol Genet 2017;26(10):1966–1978. 67. Li J, Lange LA, Duan Q, Lu Y, Singleton AB, Zonderman AB, Evans MK, Li Y, Taylor HA, Willis MS, et al. Genome-wide admixture and association study of serum iron, ferritin, transferrin saturation and total iron binding capacity in African Americans. Hum Mol Genet 2015;24(2):572–581. 68. Kamatani Y, Matsuda K, Okada Y, Kubo M, Hosono N, Daigo Y, Nakamura Y, Kamatani N. Genome-wide association study of hematological and biochemical traits in a Japanese population. Nat Genet 2010;42(3):210–215. 69. De Gobbi M, Roetto A, Piperno A, Mariani R, Alberti F, Papanikolaou G, Politou M, Lockitch G, Girelli D, Fargion S, et al. Natural history of juvenile haemochromatosis. Br J Haematol 2002;117(4):973–979. 148 70. Sandhu K, Flintoff K, Chatfield MD, Dixon JL, Ramm LE, Ramm GA, Powell LW, Nathan Subramaniam V, Wallace DF. Phenotypic analysis of hemochromatosis subtypes reveals variations in severity of iron overload and clinical disease. Blood 2018;132(1):101–110. 71. Koyama C, Hayashi H, Wakusawa S, Ueno T, Yano M, Katano Y, Goto H, Kidokoro R. Three patients with middle-age-onset hemochromatosis caused by novel mutations in the hemojuvelin gene. J Hepatol 2005;43(4):740–742. 72. Ravasi G, Pelucchi S, Mariani R, Silvestri L, Camaschella C, Piperno A. A severe hemojuvelin mutation leading to late onset of HFE2-hemochromatosis. Dig Liver Dis 2018;50(8):859–862. 73. Kong X, Xie L, Zhu H, Song L, Xing X, Yang W, Chen X. Genotypic and phenotypic spectra of hemojuvelin mutations in primary hemochromatosis patients: A systematic review. Orphanet J Rare Dis 2019;14(1):171. 74. Wallace DF, Subramaniam VN. The global prevalence of HFE and non-HFE hemochromatosis estimated from analysis of next-generation sequencing data. Genet Med 2016;18(6):618–626. 75. Papanikolaou G, Samuels ME, Ludwig EH, MacDonald MLE, Franchini PL, Dubé MP, Andres L, MacFarlane J, Sakellaropoulos N, Politou M, et al. Mutations in HFE2 cause iron overload in chromosome 1q-linked juvenile hemochromatosis. Nat Genet 2004;36(1):77–82. 76. Gehrke S, Pietrangelo A, Kaščák M, Braner A, Eisold M, Kulaksiz H, Herrmann T, Hebling U, Bents K, Gugler R, et al. HJV gene mutations in European patients with juvenile hemochromatosis. Clin Genet 2005;67(5):425–428. 149 77. McDonald CJ, Wallace DF, Crawford DHG, Subramaniam VN. Iron storage disease in Asia-Pacific populations: The importance of non- HFE mutations. J Gastroenterol Hepatol 2013;28(7):1087–1094. 78. Huang FW, Rubio-Aliaga I, Kushner JP, Andrews NC, Fleming MD. Identification of a novel mutation (C321X) in HJV. Blood 2004;104(7):2176–2177. 79. Lv T, Zhang W, Xu A, Li Y, Zhou D, Zhang B, Li X, Zhao X, Wang Y, Wang X, et al. Non-HFE mutations in haemochromatosis in China: Combination of heterozygous mutations involving HJV signal peptide variants. J Med Genet 2018;55(10):650–660. 80. Li S, Xue J, Chen B, Wang Q, Shi M, Xie X, Zhang L. Two middle-age-onset hemochromatosis patients with heterozygous mutations in the hemojuvelin gene in a Chinese family. Int J Hematol 2014;99(4):487–492. 81. Yuanfeng L, Hongxing Z, Haitao Z, Xiaobo P, Lili B, Fuchu H, Zewu Q, Gangqiao Z. Mutation analysis of the pathogenic gene in a Chinese family with hereditary hemochromatosis. Yi Chuan 2014;36(11):1152–1158. 82. Kaneko Y, Miyajima H, Piperno A, Tomosugi N, Hayashi H, Morotomi N, Tsuchida KI, Ikeda T, Ishikawa A, Ota Y, et al. Measurement of serum hepcidin-25 levels as a potential test for diagnosing hemochromatosis and related disorders. J Gastroenterol 2010;45(11):1163–1171. 83. Hattori A, Miyajima H, Tomosugi N, Tatsumi Y, Hayashi H, Wakusawa S. Clinicopathological study of Japanese patients with genetic iron overload syndromes. Pathol Int 2012;62(9):612–618. 84. Ikuta K, Hatayama M, Addo L, Toki Y, Sasaki K, Tatsumi Y, Hattori A, Kato A, Kato K, Hayashi H, et al. Iron overload patients with unknown etiology from national survey in 150 Japan. Int J Hematol 2017;105(3):353–360. 85. Nagayoshi Y, Nakayama M, Suzuki S, Hokamaki J, Shimomura H, Tsujita K, Fukuda M, Yamashita T, Nakamura Y, Sugiyama S, et al. A Q312X mutation in the hemojuvelin gene is associated with cardiomyopathy due to juvenile haemochromatosis. Eur J Heart Fail 2008;10(10):1001–1006. 86. Dhillon BK, Chopra G, Jamwal M, Chandak GR, Duseja A, Malhotra P, Chawla YK, Garewal G, Das R. Adult onset hereditary hemochromatosis is associated with a novel recurrent Hemojuvelin (HJV) gene mutation in north Indians. Blood Cells, Mol Dis 2018;73:14–21. 87. Chun YL, Merryweather-Clarke AT, Viprakasit V, Chinthammitr Y, Srichairatanakool S, Limwongse C, Oleesky D, Robins AJ, Hudson J, Wai P, et al. Iron overload in the Asian community. Blood 2009;114(1):20–25. 88. Barton JC, LaFreniere SA, Leiendecker-Foster C, Li H, Acton RT, Press RD, Eckfeldt JH. HFE, SLC40A1, HAMP, HJV, TFR2, and FTL mutations detected by denaturing high- performance liquid chromatography after iron phenotyping and HFE C282Y and H63D genotyping in 785 HEIRS Study participants. Am J Hematol 2009;84(11):710–714. 89. Lee PL, Barton JC, Brandhagen D, Beutler E. Hemojuvelin ( HJV ) mutations in persons of European, African-American and Asian ancestry with adult onset haemochromatosis. Br J Haematol 2004;127(2):224–229. 90. Hamdi-Rozé H, Ben Ali Z, Ropert M, Detivaud L, Aggoune S, Simon D, Pelletier G, Deugnier Y, David V, Bardou-Jacquet E. Variable expressivity of HJV related hemochromatosis: “Juvenile” hemochromatosis? Blood Cells, Mol Dis 2019;74:30–33. 91. Lanzara C, Roetto A, Daraio F, Rivard S, Ficarella R, Simard H, Cox TM, Cazzola M, 151 Piperno A, Gimenez-Roqueplo AP, et al. Spectrum of hemojuvelin gene mutations in 1q- linked juvenile hemochromatosis. Blood 2004;103(11):4317–4321. 92. Santos PCJL, Cançado RD, Pereira AC, Schettert IT, Soares RAG, Pagliusi RA, Hirata RDC, Hirata MH, Teixeira AC, Figueiredo MS, et al. Hereditary hemochromatosis: Mutations in genes involved in iron homeostasis in Brazilian patients. Blood Cells, Mol Dis 2011;46(4):302–307. 93. Mendes AI, Ferro A, Martins R, Picanço I, Gomes S, Cerqueira R, Correia M, Nunes AR, Esteves J, Fleming R, et al. Non-classical hereditary hemochromatosis in Portugal: Novel mutations identified in iron metabolism-related genes. Ann Hematol 2009;88(3):229–234. 94. Roettol A, Papanikolaou G, Politou M, Alberti F, Girelli D, Christakis J, Loukopoulos D, Camaschella C. Mutant antimicrobial peptide hepcidin is associated with severe juvenile hemochromatosis. Nat Genet 2003;33(1):21–22. 95. Matthes T, Aguilar-Martinez P, Pizzi-Bosman L, Darbellay R, Rubbia-Brandt L, Giostra E, Michel M, Ganz T, Beris P. Severe hemochromatosis in a Portuguese family associated with a new mutation in the 5′-UTR of the HAMP gene. Blood 2004;104(7):2181–2183. 96. Porto G, Roetto A, Daraio F, Pinto JP, Almeida S, Bacelar C, Nemeth E, Ganz T, Camaschella C. A Portuguese patient homozygous for the -25G+>A mutation of the HAMP promoter shows evidence of steady-state transcription but fails to up-regulate hepcidin levels by iron. Blood 2005;106(8):2922–2923. 97. Fonseca PFS, Cançado RD, Uellendahl Lopes MM, Correia E, Lescano MA, Santos PCJL. HAMP gene mutation associated with juvenile hemochromatosis in Brazilian patients. Acta Haematol 2016;135(4):228–231. 98. Lescano MA, Tavares LC, Santos PCJL. Juvenile hemochromatosis: HAMP mutation and 152 severe iron overload treated with phlebotomies and deferasirox . World J Clin Cases 2017;5(10):381–383. 99. Roetto A, Daraio F, Porporato P, Caruso R, Cox TM, Cazzola M, Gasparini P, Piperno A, Camaschella C. Sceening hepcidin for mutations in juvenile hemochromatosis: Identification of a new mutation (C70R). Blood 2004;103(6):2407–2409. 100. Majore S, Binni F, Pennese A, De Santis A, Crisi A, Grammatico P. HAMP gene mutation c.208T>C (p.C70R) identified in an Italian patient with severe hereditary hemochromatosis. Hum Mutat 2004;23(4):400. 101. Delatycki M, Allen K, Gow P, MacFarlane J, Radomski C, Thompson J, Hayden M, Goldberg Y, Samuels M. A homozygous HAMP mutation in a multiply consanguineous family with pseudo-dominant juvenile hemochromatosis. Clin Genet 2004;65(5):378–383. 102. Hattori A, Tomosugi N, Tatsumi Y, Suzuki A, Hayashi K, Katano Y, Inagaki Y, Ishikawa T, Hayashi H, Goto H, et al. Identification of a novel mutation in the HAMP gene that causes non-detectable hepcidin molecules in a Japanese male patient with juvenile hemochromatosis. Blood Cells, Mol Dis 2012;48(3):179–182. 103. Gichohi-Wainaina WN, Tanaka T, Towers GW, Verhoef H, Veenemans J, Talsma EF, Harryvan J, Boekschoten MV, Feskens EJ, Melse-Boonstra A. Associations between common variants in iron-related genes with haematological traits in populations of african ancestry. PLoS One 2016;11(6):e0157996. 104. Ravasi G, Rausa M, Pelucchi S, Arosio C, Greni F, Mariani R, Pelloni I, Silvestri L, Pineda P, Camaschella C, et al. Transferrin receptor 2 mutations in patients with juvenile hemochromatosis phenotype. Am J Hematol 2015;90(12):E226–E227. 105. Piperno A, Roetto A, Mariani R, Pelucchi S, Corengia C, Daraio F, Piga A, Garozzo G, 153 Camaschella C. Homozygosity for transferrin receptor-2 Y250X mutation induces early iron overload. Haematologica 2004;89(3):359-60. 106. Le Gac G, Mons F, Jacolot S, Scotet V, Ferec C, Frebourg T. Early onset hereditary hemochromatosis resulting from a novel TFR2 gene nonsense mutation (R105X) in two siblings of north French descent. Br J Haematol 2004;125(5):674–678. 107. Joshi R, Shvartsman M, Moran E, Lois S, Aranda J, Barque A, de la Cruz X, Bruguera M, Vagace JM, Gervasini G, et al. Functional consequences of transferrin receptor-2 mutations causing hereditary hemochromatosis type 3. Mol Genet Genomic Med 2015;3(3):221–232. 108. Girelli D, Bozzini C, Roetto A, Alberti F, Daraio F, Colombari R, Olivieri O, Corrocher R, Camaschella C. Clinical and pathologic findings in hemochromatosis type 3 due to a novel mutation in transferrin receptor 2 gene. Gastroenterology 2002;122(5):1295–1302. 109. Hattori A, Wakusawa S, Hayashi H, Harashima A, Sanae F, Kawanaka M, Yamada G, Yano M, Yoshioka K. AVAQ 594-597 deletion of the TfR2 gene in a Japanese family with hemochromatosis. Hepatol Res 2003;26(2):154–156. 110. Zamani F, Bagheri Z, Bayat M, Fereshtehnejad SM, Basi A, Najmabadi H, Ajdarkosh H. Iranian hereditary hemochromatosis patients: Baseline characteristics, laboratory data and gene mutations. Med Sci Monit 2012;18(10):CR622. 111. Camaschella C, Roetto A, Calì A, De Gobbi M, Garozzo G, Carella M, Majorano N, Totaro A, Gasparini P. The gene TFR2 is mutated in a new type of haemochromatosis mapping to 7q22. Nat Genet 2000;25(1):14–15. 112. Lee PL, Halloran C, West C, Beutler E. Mutation analysis of the transferrin receptor-2 gene in patients with iron overload. Blood Cells, Mol Dis 2001;27(1):285–289. 154 113. Hayashi H, Wakusawa S, Motonishi S, Miyamoto KI, Okada H, Inagaki Y, Ikeda T. Genetic background of primary iron overload syndromes in Japan. Intern Med 2006;45(20):1107–1111. 114. Soranzo N, Spector TD, Mangino M, Kühnel B, Rendon A, Teumer A, Willenborg C, Wright B, Chen L, Li M, et al. A genome-wide meta-analysis identifies 22 loci associated with eight hematological parameters in the HaemGen consortium. Nat Genet 2009;41(11):1182–1190. 115. An P, Wu Q, Wang H, Guan Y, Mu M, Liao Y, Zhou D, Song P, Wang C, Meng L, et al. TMPRSS6, but not TF, TFR2 or BMP2 variants are associated with increased risk of iron- deficiency anemia. Hum Mol Genet 2012;21(9):2124–2131. 116. Nemeth E, Tuttle MS, Powelson J, Vaughn MD, Donovan A, Ward DMV, Ganz T, Kaplan J. Hepcidin regulates cellular iron efflux by binding to ferroportin and inducing its internalization. Science 2004;306(5704):2090–2093. 117. Drakesmith H, Nemeth E, Ganz T. Ironing out Ferroportin. Cell Metab 2015;22(5):777– 787. 118. Pietrangelo A. The ferroportin disease. Blood Cells, Mol Dis 2004;32(1):131–138. 119. Mayr R, Janecke AR, Schranz M, Griffiths WJH, Vogel W, Pietrangelo A, Zoller H. Ferroportin disease: A systematic meta-analysis of clinical and molecular findings. J Hepatol 2010;53(5):941–949. 120. Vlasveld LT, Janssen R, Bardou-Jacquet E, Venselaar H, Hamdi-Roze H, Drakesmith H, Swinkels DW. Twenty years of ferroportin disease: A review or an update of published clinical, biochemical, molecular, and functional features. Pharmaceuticals 2019;12(3):132. 121. Roetto A, Merryweather-Clarke AT, Daraio F, Daraio F, Livesey K, Pointon JJ, 155 Barbabietola G, Piga A, Mackie PH, Robson KJH, Camaschella C. A valine deletion of ferroportin 1: A common mutation in hemochromatosis type 4. Blood 2002;100(2):733– 734. 122. Callebaut I, Joubrel R, Pissard S, Kannengiesser C, Gérolami V, Ged C, Cadet E, Cartault F, Ka C, Gourlaouen I, et al. Comprehensive functional annotation of 18 missense mutations found in suspected hemochromatosis type 4 patients. Hum Mol Genet 2014;23(17):4479–4490. 123. Cazzola M, Cremonesi L, Papaioannou M, Soriani N, Kioumi A, Charalambidou A, Paroni R, Romtsou K, Levi S, Ferrari M, et al. Genetic hyperferritinaemia and reticuloendothelial iron overload associated with a three base pair deletion in the coding region of the ferroportin gene (SLC11A3). Br J Haematol 2002;119(2):539–546. 124. Speletas M, Kioumi A, Loules G, Hytiroglou P, Tsitouridis J, Christakis J, Germenis AE. Analysis of SLC40A1 gene at the mRNA level reveals rapidly the causative mutations in patients with hereditary hemochromatosis type IV. Blood Cells, Mol Dis 2008;40(3):353– 359. 125. Devalia V, Carter K, Walker AP, Perkins SJ, Worwood M, May A, Dooley JS. Autosomal dominant reticuloendothelial iron overload associated with a 3-base pair deletion in the ferroportin 1 gene (SLC11A3). Blood 2002;100(2):695–697. 126. Wallace DF, Browett P, Wong P, Kua H, Ameratunga R, Subramaniam VN. Identification of ferroportin disease in the Indian subcontinent. Gut 2005;54(4):567–568. 127. Beutler E, Barton JC, Felitti VJ, Gelbart T, West C, Lee PL, Waalen J, Vulpe C. Ferroportin 1 (SCL40A1) variant associated with iron overload in African-Americans. Blood Cells, Mol Dis 2003;31(3):305–309. 156 128. Kasvosve I, Gomo ZAR, Nathoo KJ, Matibe P, Mudenge B, Loyevsky M, Gordeuk VR. Effect of ferroportin Q248H polymorphism on iron status in African children. Am J Clin Nutr 2005;82(5):1102–1106. 129. McNamara L, Gordeuk VR, MacPhail AP. Ferroportin (Q248H) mutations in African families with dietary iron overload. J Gastroenterol Hepatol 2005;20(12):1855–1858. 130. Albuquerque D, Manco L, Loua KM, Arez AP, De Jesus Trovoada M, Relvas L, Millimono TS, Rath SL, Lopes D, Nogueira F, et al. SLC40A1 Q248H allele frequencies and associated SLC40A1 haplotypes in three West African population samples. Ann Hum Biol 2011;38(3):378–381. 131. Gordeuk VR, Caleffi A, Corradini E, Ferrara F, Jones RA, Castro O, Onyekwere O, Kittles R, Pignatti E, Montosi G, et al. Iron overload in Africans and African-Americans and a common mutation in the SCL40A1 (ferroportin 1) gene. Blood Cells, Mol Dis 2003;31(3):299–304. 132. Kasvosve I, Tshwenyego U, Phuthego T, Koto G, Zachariah M, Nyepetsi NG, Motswaledi MS. Serum ferritin concentration is affected by ferroportin Q248H mutation in Africans. Clin Chim Acta 2015;444:257–259. 133. Rivers CA, Barton JC, Gordeuk VR, Acton RT, Speechley MR, Snively BM, Leiendecker-Foster C, Press RD, Adams PC, McLaren GD, et al. Association of ferroportin Q248H polymorphism with elevated levels of serum ferritin in African Americans in the Hemochromatosis and Iron Overload Screening (HEIRS) Study. Blood Cells, Mol Dis 2007;38(3):247–252. 134. Muriuki JM, Mentzer AJ, Band G, Gilchrist JJ, Carstensen T, Lule SA, Goheen MM, Joof F, Kimita W, Mogire R, et al. The ferroportin Q248H mutation protects from anemia, but 157 not malaria or bacteremia. Sci Adv 2019;5(9):eaaw0109. 135. Meidtner K, Podmore C, Kröger J, Van Der Schouw YT, Bendinelli B, Agnoli C, Arriola L, Barricarte A, Boeing H, Cross AJ, et al. Interaction of dietary and genetic factors influencing body iron status and risk of type 2 diabetes within the EPIC-InterAct study. Diabetes Care 2018;41(2):277–285. 136. Liao M, Shi J, Huang L, Gao Y, Tan A, Wu C, Lu Z, Yang X, Zhang S, Hu Y, et al. Genome-wide association study identifies variants in PMS1 associated with serum ferritin in a Chinese population. PLoS One 2014;9(8):e105844. 137. Wang CY, Meynard D, Lin HY. The role of TMPRSS6/matriptase-2 in iron regulation and anemia. Front Pharmacol 2014;5:114. 138. Finberg KE, Heeney MM, Campagna DR, Aydinok Y, Pearson HA, Hartman KR, Mayo MM, Samuel SM, Strouse JJ, Markianos K, et al. Mutations in TMPRSS6 cause iron- refractory iron deficiency anemia (IRIDA). Nat Genet 2008;40(5):569–571. 139. Du X, She E, Gelbart T, Truksa J, Lee P, Xia Y, Khovananth K, Mudd S, Mann N, Moresco EMY, et al. The serine protease TMPRSS6 is required to sense iron deficiency. Science 2008;320(5879):1088–1092. 140. Silvestri L, Guillem F, Pagani A, Nai A, Oudin C, Silva M, Toutain F, Kannengiesser C, Beaumont C, Camaschella C, et al. Molecular mechanisms of the defective hepcidin inhibition in TMPRSS6 mutations associated with iron-refractory iron deficiency anemia. Blood 2009;113(22):5605–5608. 141. Donker AE, Raymakers RAP, Thom Vlasveld L, Van Barneveld T, Terink R, Dors N, Brons PPT, Knoers NVAM, Swinkels DW. Practice guidelines for the diagnosis and management of microcytic anemias due to genetic disorders of iron metabolism or heme 158 synthesis. Blood 2014;123(25):3873–3886. 142. Donker AE, Schaap CCM, Novotny VMJ, Smeets R, Peters TMA, van den Heuvel BLP, Raphael MF, Rijneveld AW, Appel IM, Vlot AJ, et al. Iron refractory iron deficiency anemia: a heterogeneous disease that is not always iron refractory. Am J Hematol 2016;91(12):E482–E490. 143. De Falco L, Sanchez M, Silvestri L, Kannengiesser C, Muckenthaler MU, Iolascon A, Gouya L, Camaschella C, Beaumont C. Iron refractory iron deficiency anemia. Haematologica 2013;98(6):845–853. 144. de Falco L, Silvestri L, Kannengiesser C, Morán E, Oudin C, Rausa M, Bruno M, Aranda J, Argiles B, Yenicesu I, et al. Functional and clinical impact of novel Tmprss6 variants in iron-refractory iron-deficiency Anemia patients and genotype-phenotype studies. Hum Mutat 2014;35(11):1321–1329. 145. Benyamin B, Ferreira MAR, Willemsen G, Gordon S, Middelberg RPS, McEvoy BP, Hottenga JJ, Henders AK, Campbell MJ, Wallace L, et al. Common variants in TMPRSS6 are associated with iron status and erythrocyte volume. Nat Genet 2009;41(11):1173– 1175. 146. Tanaka T, Roy CN, Yao W, Matteini A, Semba RD, Arking D, Walston JD, Fried LP, Singleton A, Guralnik J, et al. A genome-wide association analysis of serum iron concentrations. Blood 2010;115(1):94–96. 147. Chambers JC, Zhang W, Li Y, Sehmi J, Wass MN, Zabaneh D, Hoggart C, Bayele H, McCarthy MI, Peltonen L, et al. Genome-wide association study identifies variants in TMPRSS6 associated with hemoglobin levels. Nat Genet 2009;41(11):1170–1172. 148. Wilman HR, Parisinos CA, Atabaki-Pasdar N, Kelly M, Thomas EL, Neubauer S, 159 Jennison C, Ehrhardt B, Baum P, Schoelsch C, et al. Genetic studies of abdominal MRI data identify genes regulating hepcidin as major determinants of liver iron concentration. J Hepatol 2019;71(3):594–602. 149. Kloss-Brandstätter A, Erhart G, Lamina C, Meister B, Haun M, Coassin S, Seifert M, Klein-Franke A, Paulweber B, Kedenko L, et al. Candidate Gene Sequencing of SLC11A2 and TMPRSS6 in a Family with Severe Anaemia: Common SNPs, Rare Haplotypes, No Causative Mutation. PLoS One 2012;7(4):e35015. 150. He M, Workalemahu T, Manson JAE, Hu FB, Qi L. Genetic determinants for body iron store and type 2 diabetes risk in US men and women. PLoS One. 2012;7(7):e40919. 151. Gichohi-Wainaina WN, Towers GW, Swinkels DW, Zimmermann MB, Feskens EJ, Melse-Boonstra A. Inter-ethnic differences in genetic variants within the transmembrane protease, serine 6 (TMPRSS6) gene associated with iron status indicators: a systematic review with meta-analyses. Genes Nutr 2015;10(1):442. 152. Jallow MW, Cerami C, Clark TG, Prentice AM, Campino S. Differences in the frequency of genetic variants associated with iron imbalance among global populations. PLoS One 2020;15(7):e0235141. 153. Goldwurm S, Casati C, Venturi N, Strada S, Santambrogio P, Indraccolo S, Arosio P, Cazzola M, Piperno A, Masera G, et al. Biochemical and genetic defects underlying human congenital hypotransferrinemia. Hematol J 2000;1(6):390–398. 154. Trombini P, Coliva T, Nemeth E, Mariani R, Ganz T, Biondi A, Piperno A. Effects of plasma transfusion on hepcidin production in human congenital hypotransferrinemia. Haematologica 2007;92(10):1407–1410. 155. Nazir H, Al-Khabori MK, Kohistani N, Puchalka J, Klein C. A Rare Case of Congenital 160 Atransferrinemia: International Collaboration for Genetic Diagnosis. Blood 2014;124(21):4883–4883. 156. Dabboubi R, Amri Y, Yahyaoui S, Mahjoub R, Sahli CA, Sahli C, Hadj Fredj S, Bibi A, Sammoud A, Messaoud T. A new case of congenital atransferrinemia with a novel splice site mutation: c.293-63del. Eur J Med Genet 2020;63(5):103874. 157. McLaren CE, McLachlan S, Garner CP, Vulpe CD, Gordeuk VR, Eckfeldt JH, Adams PC, Acton RT, Murray JA, Leiendecker-Foster C, et al. Associations between single nucleotide polymorphisms in iron-related genes and iron status in multiethnic populations. PLoS One 2012;7(6):e38339. 158. Kayaalti Z, Akyüzlü DK, Söylemezoğlu T. Evaluation of the effect of divalent metal transporter 1 gene polymorphism on blood iron, lead and cadmium levels. Environ Res 2015;137:8–13. 159. Tolone C, Bellini G, Punzo F, Papparella A, Miele E, Vitale A, Nobili B, Strisciuglio C, Rossi F. The DMT1 IVS4+44C>A polymorphism and the risk of iron deficiency anemia in children with celiac disease. PLoS One 2017;12(10):e0185822. 160. Constantine CC, Anderson GJ, Vulpe CD, McLaren CE, Bahlo M, Yeap HL, Gertig DM, Osborne NJ, Bertalli NA, Beckman KB, et al. A novel association between a SNP in CYBRD1 and serum ferritin levels in a cohort study of HFE hereditary haemochromatosis. Br J Haematol 2009;147(1):140–149. 161. Pelucchi S, Mariani R, Calza S, Fracanzani AL, Modignani GL, Bertola F, Busti F, Trombini P, Fraquelli M, Forni GL, et al. CYBRD1 as a modifier gene that modulates iron phenotype in HFE p.C282Y homozygous patients. Haematologica 2012;97(12):1818– 1825. 161 162. Milet J, Déhais V, Bourgain C, Jouanolle AM, Mosser A, Perrin M, Morcet J, Brissot P, David V, Deugnier Y, et al. Common variants in the BMP2, BMP4, and HJV genes of the hepcidin regulation pathway modulate HFE hemochromatosis penetrance. Am J Hum Genet 2007;81(4):799–807. 163. Gill D, Benyamin B, Moore LSP, Monori G, Zhou A, Koskeridis F, Evangelou E, Laffan M, Walker AP, Tsilidis KK, et al. Associations of genetically determined iron status across the phenome: A mendelian randomization study. PLOS Med 2019;16(6):e1002833. 164. Wolf I, Jenkins BJ, Liu Y, Seiffert M, Custodio JM, Young P, Rohrschneider LR. Gab3, a New DOS/Gab Family Member, Facilitates Macrophage Differentiation. Mol Cell Biol 2002;22(1):231–244. 165. Wöhrle FU, Daly RJ, Brummer T. Function, regulation and pathological roles of the Gab/DOS docking proteins. Cell Commun Signal 2009;7(1):1–28. 166. Hirano N, Butler MO, Guinan EC, Nadler LM, Kojima S. Presence of anti-kinectin and anti-PMS1 antibodies in Japanese aplastic anaemia patients. Br J Haematol 2005;128(2):221–223. 167. James WPT, Johnson RJ, Speakman JR, Wallace DC, Frühbeck G, Iversen PO, Stover PJ. Nutrition and its role in human evolution. J Intern Med 2019;285(5):533–549. 168. McLaren CE, Barton JC, Adams PC, Harris EL, Acton RT, Press N, Reboussin DM, McLaren GD, Sholinsky P, Walker AP, et al. Hemochromatosis and iron overload screening (HEIRS) Study design for an evaluation of 100,000 primary care-based adults. Am J Med Sci 2003;325(2):53–62. 169. Endres-Dighe SM, Guo Y, Kanias T, Lanteri M, Stone M, Spencer B, Cable RG, Kiss JE, Kleinman S, Gladwin MT, et al. Blood, sweat, and tears: Red Blood Cell-Omics study 162 objectives, design, and recruitment activities. Transfusion 2019;59(1):46–56. 170. Zacharski LR, Shamayeva G, Chow BK, Depalma RG. Racial differences in iron measures andoutcomes observed during an iron reduction trial in peripheral arterial disease. J Health Care Poor Underserved 2015;26(1):243–259. 171. Beutler E, West C. Hematologic differences between African-Americans and whites: The roles of iron deficiency and α-thalassemia on hemoglobin levels and mean corpuscular volume. Blood 2005;106(2):740–745. 172. Beutler E, Felitti V, Gelbart T, Ho N. The effect of HFE genotypes on measurements of iron overload in patients attending a health appraisal clinic. Ann Intern Med 2000;133(5):329-37. 173. Zacharski LR, Ornstein DL, Woloshin S, Schwartz LM. Association of age, sex, and race with body iron stores in adults: Analysis of NHANES III data. Am Heart J 2000;40(1):98- 104. 174. Dekker LH, Nicolaou M, Van Der A DL, Busschers WB, Brewster LM, Snijder MB, Stronks K, Van Valkengoed IGM. Sex differences in the association between serum ferritin and fasting glucose in type 2 diabetes among South Asian Surinamese, African Surinamese, and ethnic Dutch: The population-based SUNSET study. Diabetes Care 2013;36(4):965–971. 175. Jenum AK, Sletner L, Voldner N, Vangen S, Mørkrid K, Andersen LF, Nakstad B, Skrivarhaug T, Rognerud-jensen O, Roald B, et al. The STORK Groruddalen research programme: A population-based cohort study of gestational diabetes, physical activity, and obesity in pregnancy in a multiethnic population. Rationale, methods, study population, and participation rates. Scand J Public Health 2010;38(5):60–70. 163 176. Godsland IF, Seed M, Simpson R, Broom G, Wynn V. Comparison of haematological indices between women of four ethnic groups and the effect of oral contraceptives. J Clin Pathol 1983;36(2):184–191. 177. Næss-Andresen ML, Eggemoen ÅR, Berg JP, Falk RS, Jenum AK. Serum ferritin, soluble transferrin receptor, and total body iron for the detection of iron deficiency in early pregnancy: A multiethnic population-based study with low use of iron supplements. Am J Clin Nutr 2019;109(3):566-575. 178. Spencer BR, Guo Y, Cable RG, Kiss JE, Busch MP, Page GP, Endres-Dighe SM, Kleinman S, Glynn SA, Mast AE. Iron status and risk factors for iron depletion in a racially/ethnically diverse blood donor population. Transfusion 2019;59(10):3146–3156. 179. Barton JC, Acton RT, Dawkins FW, Adams PC, Lovato L, Leiendecker-Foster C, McLaren CE, Reboussin DM, Speechley MR, Gordeuk VR, et al. Initial screening transferrin saturation values, serum ferritin concentrations, and HFE genotypes in whites and blacks in the Hemochromatosis and Iron Overload Screening Study. Genet Test 2005;9(3):231-41. 180. Miller EM. Iron status and reproduction in US women: National health and nutrition examination survey, 1999-2006. PLoS One 2014;9(11):e112216. 181. Dallman PR, Barr GD, Allen CM, Shinefield HR. Hemoglobin concentration in white, black, and Oriental children: is there a need for separate criteria in screening for anemia? Am J Clin Nutr 1978;31(3):377-80. 182. Perry GS, Byers T, Yip R, Margen S. Iron nutrition does not account for the hemoglobin differences between blacks and whites. J Nutr 1992;122(7):1417–1424. 183. McLaren CE, Li KT, Gordeuk VR, Hasselblad V, McLaren GD. Relationship between 164 transferrin saturation and iron stores in the African American and US Caucasian populations: Analysis of data from the third National Health and Nutrition Examination Survey. Blood 2001;98(8):2345–2351. 184. Johnson-Spear MA, Yip R. Hemoglobin difference between black and white women with comparable iron status: justification for race-specific anemia criteria1. Am J Clin Nutr 1994;60(1):117-21. 185. Meyers LD, Habicht JP, Johnson CL. Components of the difference in hemoglobin concentrations in blood between black and white women in the united states. Am J Epidemiol 1979;109(5):539–549. 186. World Health Organization, Centers for Disease Control and Prevention. Assessing the Iron StatuS of populations Second edition Including Literature Reviews Geneva, Switzerland: World Health Organization, 2007. 187. Pan Y, Jackson RT. Ethnic difference in the relationship between acute inflammation and serum ferritin in US adult males. Epidemiol Infect 2008;136(3):421–431. 188. Zacharski LR, Shamayeva G, Chow BK, DePalma RG. Racial health disparities, and variant red cell and iron homeostasis. J Health Care Poor Underserved 2016;27(2):741– 761. 189. Pan Y, Jackson RT. Insights into the ethnic differences in serum ferritin between black and white US adult men. Am J Hum Biol 2008;20(4):406–416. 190. Barton JC, Wiener HH, Acton RT, Adams PC, Eckfeldt JH, Gordeuk VR, Harris EL, McLaren CE, Harrison H, McLaren GD, et al. Prevalence of iron deficiency in 62,685 women of seven race/ethnicity groups: The HEIRS Study. PLoS One 2020;15(4):e0232125. 165 191. Gordeuk VR, Brannon PM. Ethnic and genetic factors of iron status in women of reproductive age. Am J Clin Nutr 2017;106(Suppl 6):1594S-1599S. 192. Cossrow N, Falkner B. Race/Ethnic Issues in Obesity and Obesity-Related Comorbidities. J Clin Endocrinol Metab 2004;89(6):2590–2594. 193. Harris EL, McLaren CE, Reboussin DM, Gordeuk VR, Barton JC, Acton RT, McLaren GD, Vogt TM, Snively BM, Leiendecker-Foster C, et al. Serum Ferritin and Transferrin Saturation in Asians and Pacific Islanders. Arch Intern Med 2007;167(7):722. 194. Oh SH, Jeong TD, Lee W, Chun S, Min WK. Analysis of Familial Tendencies in Transferrin Saturation in a Korean Population. Dig Dis Sci 2015;60(10):3136–3141. 195. Lee SH, Kim JW, Shin SH, Kang KP, Choi HC, Choi SH, Park KU, Kim HY, Kang W, Jeong SH. HFE gene mutations, serum ferritin level, transferrin saturation, and their clinical correlates in a Korean population. Dig Dis Sci 2009;54(4):879–886. 196. Hsu WC, Araneta MRG, Kanaya AM, Chiang JL, Fujimoto W. BMI cut points to identify at-risk Asian Americans for type 2 diabetes screening. Diabetes Care 2015;38(1):150–8. 197. WHO Expert Consultation. Appropriate body-mass index for Asian populations and its implications for policy and intervention strategies. Lancet 2004;363(9403):157–163. 198. Barton JC, Acton RT, Lovato L, Speechley MR, McLaren CE, Harris EL, Reboussin DM, Adams PC, Dawkins FW, Gordeuk VR, et al. Initial screening transferrin saturation values, serum ferritin concentrations, and HFE genotypes in Native Americans and whites in the Hemochromatosis and Iron Overload Screening Study. Clin Genet 2006;69(1):48– 57. 199. Walters GO, Miller FM, Worwood M. Serum ferritin concentration and iron stores in normal subjects. J Clin Pathol 1973;26(10):770–772. 166 200. Lynch S. The rationale for selecting and standardizing iron status indicators. Geneva, World Health Organization, 2012. 201. O’Brien KO, Ru Y. Iron status of North American pregnant women: An update on longitudinal data and gaps in knowledge from the United States and Canada. Am J Clin Nutr 2017;106(Suppl 6):1647S-16 167 CHAPTER 5 SUMMARY AND CONCLUSIONS 168 Summary This doctoral research investigated three major factors that influence iron homeostasis, namely physiological, environmental, and genetic factors. Each of these factors was a specific aim of the dissertation. To address a physiological determinant of iron homeostasis, this research provided a large normative analysis of determinants of maternal Hb and identified factors associated with risk of maternal anemia and elevated maternal Hb concentrations in healthy U.S. pregnant women. To address an environmental situation that impacts iron homeostasis, the impact of parasitic infections and iron supplementation on iron absorption was investigated and identified variables associated with the change in RBC iron isotope enrichment over a 3-month period was evaluated among a group of toddlers residing in a malaria and helminth endemic environment. As for the genetic control of iron homeostasis, this doctoral dissertation summarized published literature on genetic determinants of iron status, characterized iron status in different ethnic populations, and statistically evaluated differences in iron status as a function of ethnicity using published data. These three specific aims were presented in Chapters Two through Four. Below is a brief summary of the conclusions from each specific aim, followed by implications and suggestions for future research. The data from the specific aim one provided one of the largest normative longitudinal analyses of Hb concentrations across pregnancy in otherwise healthy U.S. pregnant women receiving routine prenatal care. Findings from Aim 1 indicated that anemia was prevalent in otherwise healthy U.S. pregnant women receiving standard prenatal care even with the current recommendation of universal iron supplementation of U.S. pregnant women. The prevalence of anemia increased as pregnancy progressed and by late pregnancy one-quarter of women evaluated had anemia. We reported that the current CDC anemia cutoffs (that were based on the 169 5th percentile of a 1970’s reference population) may overestimate the prevalence of anemia in late pregnancy based on our more recent analysis of a much larger and more racially diverse population. Risk factors for anemia were identified as Black race or Hispanic ethnicity, being underweight at entry into pregnancy, being multiparous or carrying a multiple birth pregnancy. Identifying these risk factors may help inform targeted iron supplementation recommendations in obstetric clinical practice. These data have contributed to the current knowledge of hemoglobin concentrations in U.S. pregnant women and identified additional risk factors associated with increased risk of anemia across pregnancy. The main findings from the Aim 2 provided a better understanding of the impact of parasitic infections and iron supplementation on iron metabolism in toddlers residing in a malaria and helminth endemic environment. This study was one of the largest stable iron isotope studies and it uniquely followed the change in RBC iron isotopic enrichment over an 84-day period post- dosing in otherwise healthy toddlers residing in a malaria- and helminth-endemic area. The strongest determinant of RBC iron incorporation was TBI. TBI likely better captures iron demands as it is calculated using biomarkers of tissue demand and storage iron. We also reported that toddlers who received iron supplementation exhibited significantly greater loss of RBC iron isotope enrichment over time. This greater loss of RBC iron enrichment may be indicative of increase in size of RBC pool and increased erythropoietic demands due to the provision of iron among anemic or helminth infected toddlers. The findings from Aim 3 highlighted the shared and unique genetic variants that have been associated with iron status biomarkers and explored possible differences in iron status as a function of ethnicity using published data. For the first time, we pooled existing published data and demonstrated that East Asians had significantly higher iron status (serum ferritin, transferrin 170 saturation) compared to Europeans, African Americans, or South Asians. Supportive of other published data we also found that Black adults exhibited significantly lower Hb concentrations compared to other ethnic groups. These data contribute to this literature and may help inform subsequent population-specific recommendations and personalized nutrition interventions for iron-related disorders. Future Directions This doctoral research provided novel insights into physiological, environmental and genetic factors that impact iron homeostasis. These data also lead to new research questions. Further studies are needed to evaluate the impact of maternal anemia on iron status in the newborn at birth and to characterize its potential impact on birth outcomes. Targeted identification of women at greater risk for ID and IDA may help inform interventions to improve the in-utero environment in support of healthy birth outcomes. Additionally, for environmental factors, future longer-term stable iron isotope studies may provide greater insight into whole body iron homeostasis and adaptations to parasitic infections or iron supplementation in vulnerable age groups such as children. Lastly, more genome-wide and candidate gene association studies of iron homeostasis are needed in ethnically diverse populations. Additional characterization of the genetic basis of ethnic differences in iron metabolism and on how this affects disease susceptibility among different ethnic groups are needed to help inform population-specific dietary iron intake recommendations and to help screen at-risk populations. Multidisciplinary approaches that consider physiological, environmental and genetic factors may best predict iron disorders in specific populations and help develop the most effective interventions to maintain adequate iron status. 171 APPENDIX 172 Appendix Table 1. Commonly utilized biochemical indicators of iron status1 Biomarker Function of Indicator Confounders Cut-offs in Non-pregnant Healthy Adults SF Indicator of storage iron Inflammation and Deficiency: infection; menstruation <12-15 μg/L(1) and childbirth; blood donations; alcohol intake; Overload: liver disease and > 150 μg/L in menstruating malignancy females > 200 μg/L in males and non-menstruating females (1) sTfR Indicator of tissue iron availability Rate of erythropoiesis; Deficiency: chronic disease; sTfR >8.5 mg/L if utilizing assay dependent Ramco assay (2), or >4.4 (Ramco/other ELISA kits mg/L for women and >5.0 or Roche) mg/L for men if utilizing the Roche Hitachi analyzer (3) Overload: No set cut-off ZPP or EP Indicator of iron supply to bone Rate of erythropoiesis; Deficiency: marrow lead poisoning; low >40-80 μg EP/dL of red cells diagnostic sensitivity and (4) specificity; high day-to- day variation Overload: No set cut-off SI Circulating transferrin-bound iron Infection and Deficiency: inflammation; varies from <50-60 μg/dL (4) day-to-day and after meals; under circadian Overload: No set cut-off rhythmicity; sample easily contaminated TIBC Measure of the total number of Large overlap between No set cut-offs. TIBC by transferrin binding sites per unit normal values and values itself is not used as a volume of plasma or serum in iron deficiency measure of iron status TSAT Indicator of iron supply to tissues Same as serum iron and Deficiency: TIBC <15% (4) (Formula = serum iron/TIBC) Overload: >45% (4) 173 Appendix Table 1 (Continued) Hb Measure of anemia and proxy of Low Deficiency: functional iron sensitivity/specificity; 12 g/dL in women does not detect recent 13 g/dL in men (5) changes Overload: No set cut-off Soluble transferrin Represents iron status in the sTfR is assay dependent Deficiency: No set cut-off receptor to serum setting of inflammation, infection ferritin ratio or chronic disease Overload: No set cut-off (Formula = sTfR/SF) TBI Provides a measure of the Same as sTfR and SF Deficiency: magnitude of iron deficiency that <0 mg/kg body weight (6) is independent of the hemoglobin concentration. Helps distinguish Overload: No set cut-off between iron deficiency anemia and anemia of chronic disease (Formula = - [log (serum transferrin receptor/serum ferritin) – 2.8229]/0.1207) 1EP, erythrocyte protoporphyrin; Hb, hemoglobin; MCH, mean cell hemoglobin; MCV, mean cell volume; SF, serum ferritin; SI, serum iron; TBI; total body iron; TF, transferrin; TIBC, total iron binding capacity; TSAT, transferrin saturation; UIBC, unsaturated iron binding capacity; ZPP, zinc protoporphyrin. 174 Appendix Table 2. Genetic variants shown to be associated with biomarkers of iron status in healthy populations stratified by ethnic group1,2 Chromosome Gene(s)3 Variants4 Iron Status-Related References (amino acid change) Phenotype Europeans 6 HFE5 rs1800562 (C282Y), SF, SI, TF, TIBC, (7,8,17,9–16) rs1799945 (H63D) TSAT, SF, sTfR, sTfR:SF, UIBC 3 TF5 rs8177240, rs8177179, SI, TF, TSAT, TIBC, (7–9,11–13, rs1799852 (L247L), UIBC 15,16) rs3811647, rs1358024, rs452586, rs4428180, rs8177224, rs3811658, rs8177248, rs1880669, rs1525892, rs9824452, rs7638018, rs2280673 22 TMPRSS65 rs855791 (V736A), sTfR, sTfR:SF, SI, (7–15,18) rs4820268 (D521D), TF, TSAT, SF rs228918, rs228919, rs228921, rs5756520, rs2111833, rs2235324, rs1421312, rs2743825 2 FPN15/ rs744653 TF, SF (9,10) WDR75 3 TFR1 rs9990333 TF (9) (TFRC)5 7 TFR25 rs7385804 SI, TSAT (8,9) 6 SLC17A1 rs17342717 SF (8) 11 PCSK7 rs236918 SF (10,14) 9 ABO rs651007 SF (9) 17 TEX14 rs411988 SF (9) 11 ARNTL rs6486121 TF (9) 11 FADS1 rs174577 TF (9) 8 NAT2 rs4921915 TF (9) 3 SRPRB rs9843728, rs1830084, TIBC, UIBC (12) rs6794676, rs6794945, rs13061203, rs9853615 18 KIAA1468 rs9948708, rs11877669 TIBC, SI (12,18) 20 TMEM90B rs11700002 TIBC (12) 7 RPA3 rs10263415 sTfR (12) 8 SAMD12 rs2460970, rs880034 TBI; SF (12) 1 SLC44A3 rs735937 SI (18) 175 Appendix Table 2 (Continued) 2 CAPN13 rs17010141 SI (18) 3 GRM7 rs11131063, rs1353828, SI (18) rs9880404, rs965170, rs10866078 5 TGFBI rs756462 SI (18) 6 CD109 rs9442947 SI (18) 7 MAD1L1 rs1176717 SI (18) 8 LPL rs270 SI (18) 9 STOM rs16910559 SI (18) 11 NELL1 rs1670646 SI (18) 14 TTC7B rs1286474 SI (18) 15 SHC4 rs16961946 SI (18) 18 DLGAP1 rs17725070 SI (18) 18 PTPRM rs7505939 SI (18) Asians 22 TMPRSS65 rs855791 (V736A), SI, TSAT, FEP, SF (19,20) rs4820268 (D521D), rs2413450 3 TF5 rs3811647 TF, TIBC (19) 7 TFR25 rs7385804 SI (19) 2 PMS1/ rs574933, rs3791770, SF (21) FPN15 rs3791773 2 ANKAR rs1550388, rs1225101 SF (21) 2 OSGEPL1 rs1898569, rs4666783 SF (21) 2 ORMDL1 rs2289404 SF (21) Africans 3 TF5 rs8177253 TIBC (22) X G6PD/ rs141555380 SF (22) GAB3 6 HDGFL1/ rs115923437 TIBC (22) NRSN1 16 MAF/ rs16951289 TIBC (22) DYNLRB2 10 CUBN rs10904850 SI (12) Hispanics 6 HFE5 rs1800562 (C282Y), TIBC, TSAT, SF, SI (23) rs1799945 (H63D) 3 TF5 rs4637289, rs2692666, TIBC, TSAT (23) rs1405023, rs6762719 22 TMPRSS65 rs855791 (V736A) TSAT, SI (23) 8 PPP1R3B rs4841132 TIBC (23) 176 Appendix Table 2 (Continued) 11 FADS2/ rs174529 TIBC (23) MYRF X GAB3 rs141555380 SF (23) 1ABO, alpha 1-3-N-acetylgalactosaminyltransferase and alpha 1-3-galactosyltransferase; ANKAR, ankyrin and armadillo repeat containing ; ARNTL, aryl hydrocarbon receptor nuclear translocator like; CAPN13, calpain 13; CD109, cluster of differentiation 109; CUBN, cubilin; DLGAP1, DLG associated protein 4; DYNLRB2, dynein light chain roadblock-type 2; FADS1, fatty acid desaturase 2; FPN1, ferroportin 1; G6P, glucose-6-phosphate dehydrogenase; GAB3, growth factor receptor bound protein 2-associated protein 3; GRM7, glutamate metabotropic receptor 7; HDGFL1, hepatoma-derived growth factor-like protein 1; HFE, high iron protein; KIAA1468, RAB11 binding and LisH domain, coiled-coil and HEAT repeat containing; LPL, lipoprotein lipase; MAD, MAX dimerization protein 1; MAD1L1, mitotic arrest deficient 1 like 1; NAT2, N-acetyltransferase 2; NELL1, neural EGFL like 1; NRSN1, neurensin 1; ORMDL1, ORMDL sphingolipid biosynthesis regulator 1; OSGEPL1, O-sialoglycoprotein endopeptidase like 1; PCSK7, protein convertase suntilisin/kexin type 7; PMS1, postmeiotic segregation increased 1; PPP1R3B, protein phosphatase 1 regulatory subunit 3B; PTPRM, protein tyrosine phosphatase receptor type M; RPA3, replication protein A3; SAMD12, sterile alpha motif domain containing 12; SF, serum ferritin; SHC4, SHC adaptor protein 4; SI, serum iron; SLC44A3, solute carrier family 44 member 3; STOM, stomatin; SLC17A1, solute carrier family 17 member 1; SRPRB, signal recognition particle receptor subunit beta; TBI, total body iron; TEX14, testis expressed 14; TF, transferrin; TF, transferrin; TFR1, transferrin receptor 1; TFR2, transferrin receptor 2; TGFB1, transforming growth factor beta 1; TIBC, total iron binding capacity; TMEM90B, synapse differentiation inducing 1; TMPRSS6, transmembrane serine protease 6; TSAT, transferrin saturation;TTC7B, tetratricopeptide repeat domain 7B; UIBC, unsaturated iron binding capacity; WDR75, WD repeat domain 75. 2Both genome-wide and candidate gene association studies included. All studies adjusted for varying covariates in their analyzes (i.e. sex, age, principal component scores, among other study-specific variables). 3 Variants within or in close proximity to specified gene(s). 4 Variants shown to be associated with variation in one or more of the iron status indicators listed on the iron status-related phenotypes column. Only associations that remained significant after multiple testing correction are included. Only SNPs with known genes nearby included. 5 Known iron-related genes. 177 Appendix Table 3. Summary of literature on iron status in different ethnic populations1 Literature with Literature with Ethnic groups comparison clinical correlates characterization data data European vs. African American vs. Asian vs. Hispanic (24),(12,25–27)2,(28)3 (29)2,(30)4 European vs. African American vs. Studies designed to (31)2,(32)3 - Hispanic evaluate ethnicities European vs. African American vs. differences in iron (33) (34) Asian status European vs. African American (35–39), (44)2 (40,41)2,(42,43)5 European vs. Asian (45),(46)5 - European vs. Native American (47) - (48,49,58–62, 50– European - Studies 57),(63)3 characterizing iron (20,72–81), Asian (64–67),(21,68–71)2 status in single (82,83)2,(84)6 ethnicity African American (22) - Hispanic - (23) 1The study type was categorized as either a study characterizing iron status in large cohorts or a study designed to compare case-control to a specific clinical condition. The most common clinical conditions that might be associated with iron status examined in the literature are iron deficiency, anemia, iron overload, diabetes, and metabolic syndrome. 2Literature was excluded from statistical analyses because of repeat study cohort. 3Literature was excluded from statistical analyses because pregnant and/or breastfeeding women were the study subjects. 4Literature was excluded from statistical analyses because frequent blood donors were the study subjects. 5Literature was excluded from statistical analyses because iron status data were not available. 6Literature was excluded from statistical analyses because there was no iron status reported by sex. 178 REFERENCES 1. World Health Organization. WHO guideline on use of ferritin concentrations to assess iron status in individuals and populations. Geneva, Switzerland: World Health Organization, 2020. 2. Pfeiffer CM, Cook JD, Mei Z, Cogswell ME, Looker AC, Lacher DA. Evaluation of an automated soluble transferrin receptor (sTfR) assay on the Roche Hitachi analyzer and its comparison to two ELISA assays. Clin Chim Acta 2007;382(1– 2):112–116. 3. Kolbe-Busch S, Lotz J, Hafner G, Blanckaert NJC, Claeys G, Togni G, Carlsen J, Röddiger R, Thomas L. Multicenter evaluation of a fully mechanized soluble transferrin receptor assay on the Hitachi and Cobas Integra analyzers. The determination of reference ranges. Clin Chem Lab Med 2002;40(5):529–536. 4. World Health Organization, Centers for Disease Control and Prevention. Assessing the Iron StatuS of populations Second edition Including Literature Reviews Geneva, Switzerland: World Health Organization, 2007. 5. World Health Organization. WHO | Iron deficiency anaemia: assessment, prevention and control. Geneva, Switzerland: World Health Organization, 2001. 6. Cook JD, Flowers CH, Skikne BS. The quantitative assessment of body iron. Blood 2003; 101(9):3359-64. 7. Benyamin B, Ferreira MAR, Willemsen G, Gordon S, Middelberg RPS, McEvoy BP, Hottenga JJ, Henders AK, Campbell MJ, Wallace L, et al. Common variants in TMPRSS6 are associated with iron status and erythrocyte volume. Nat Genet 2009;41:1173–5. 179 8. Pichler I, Minelli C, Sanna S, Tanaka T, Schwienbacher C, Naitza S, Porcu E, Pattaro C, Busonero F, Zanon A, et al. Identification of a common variant in the TFR2 gene implicated in the physiological regulation of serum iron levels. Hum Mol Genet 2011;20:1232–40. 9. Beutler E, Felitti V, Gelbart T, Waalen J. Haematological effects of the C282Y HFE mutation in homozygous and heterozygous states among subjects of northern and southern European ancestry. Br J Haematol 2003;120:887–93. 10. Benyamin B, Esko T, Ried JS, Radhakrishnan A, Vermeulen SH, Traglia M, Gögele M, Anderson D, Broer L, Podmore C, et al. Novel loci affecting iron homeostasis and their effects in individuals at risk for hemochromatosis. Nat Commun 2014;5:4926. 11. Meidtner K, Podmore C, Kröger J, Van Der Schouw YT, Bendinelli B, Agnoli C, Arriola L, Barricarte A, Boeing H, Cross AJ, et al. Interaction of dietary and genetic factors influencing body iron status and risk of type 2 diabetes within the EPIC-InterAct study. Diabetes Care 2018;41:277–85. 12. He M, Workalemahu T, Manson JAE, Hu FB, Qi L. Genetic determinants for body iron store and type 2 diabetes risk in US men and women. PLoS One 2012;7(7):e40919. 13. McLaren CE, McLachlan S, Garner CP, Vulpe CD, Gordeuk VR, Eckfeldt JH, Adams PC, Acton RT, Murray JA, Leiendecker-Foster C, et al. Associations between single nucleotide polymorphisms in iron-related genes and iron status in multiethnic populations. PLoS One 2012;7(6):e38339.. 180 14. Traglia M, Girelli D, Biino G, Campostrini N, Corbella M, Sala C, Masciullo C, Viganò F, Buetti I, Pistis G, et al. Association of HFE and TMPRSS6 genetic variants with iron and erythrocyte parameters is only in part dependent on serum hepcidin concentrations. J Med Genet 2011;48:629–34. 15. Oexle K, Ried JS, Hicks AA, Tanaka T, Hayward C, Bruegel M, Gögele M, Lichtner P, Müller-Myhsok B, Döring A, et al. Novel association to the proprotein convertase PCSK7 gene locus revealed by analysing soluble transferrin receptor (sTfR) levels. Hum Mol Genet 2011;20:1042–7. 16. Benyamin B, McRae AF, Zhu G, Gordon S, Henders AK, Palotie A, Peltonen L, Martin NG, Montgomery GW, Whitfield JB, et al. Variants in TF and HFE Explain ∼40% of Genetic Variation in Serum-Transferrin Levels. Am J Hum Genet 2009;84(1): 60–65. 17. McLaren CE, Garner CP, Constantine CC, McLachlan S, Vulpe CD, Snively BM, Gordeuk VR, Nickerson DA, Cook JD, Leiendecker-Foster C, et al. Genome-wide association study identifies genetic loci associated with iron deficiency. PLoS One 2011;6:17390. 18. Tanaka T, Roy CN, Yao W, Matteini A, Semba RD, Arking D, Walston JD, Fried LP, Singleton A, Guralnik J, et al. A genome-wide association analysis of serum iron concentrations. Blood 2010;115:94–6. 19. An P, Wu Q, Wang H, Guan Y, Mu M, Liao Y, Zhou D, Song P, Wang C, Meng L, et al. TMPRSS6, but not TF, TFR2 or BMP2 variants are associated with increased risk of iron-deficiency anemia. Hum Mol Genet 2012;21:2124–31. 181 20. Gan W, Guan Y, Wu Q, An P, Zhu J, Lu L, Jing L, Yu Y, Ruan S, Xie D, et al. Association of TMPRSS6 polymorphisms with ferritin, hemoglobin, and type 2 diabetes risk in a Chinese Han population. Am J Clin Nutr 2012;95(3):626–632. 21. Liao M, Shi J, Huang L, Gao Y, Tan A, Wu C, Lu Z, Yang X, Zhang S, Hu Y, et al. Genome-wide association study identifies variants in PMS1 associated with serum ferritin in a Chinese population. PLoS One 2014;9(8):e105844. 22. Li J, Lange LA, Duan Q, Lu Y, Singleton AB, Zonderman AB, Evans MK, Li Y, Taylor HA, Willis MS, et al. Genome-wide admixture and association study of serum iron, ferritin, transferrin saturation and total iron binding capacity in African Americans. Hum Mol Genet 2015;24:572–81. 23. Raffield LM, Louie T, Sofer T, Jain D, Ipp E, Taylor KD, Papanicolaou GJ, Avilés-Santa L, Lange LA, Laurie CC, et al. Genome-wide association study of iron traits and relation to diabetes in the Hispanic Community Health Study/Study of Latinos (HCHS/SOL): Potential genomic intersection of iron and glucose regulation? Hum Mol Genet 2017;26:1966–78. 24. Beutler E, Felitti V, Gelbart T, Ho N. The effect of HFE genotypes on measurements of iron overload in patients attending a health appraisal clinic. Ann Intern Med 2000;133(5):329-37. 25. Gordeuk VR, Reboussin DM, McLaren CE, Barton JC, Acton RT, McLaren GD, Harris EL, Reiss JA, Adams PC, Speechley M, et al. Serum ferritin concentrations and body iron stores in a multicenter, multiethnic primary-care population. Am J Hematol 2008;83:618–26. 182 26. Gordeuk VR, Brannon PM. Ethnic and genetic factors of iron status in women of reproductive age. Am J Clin Nutr 2017;106(Suppl 6):1594S-1599S. 27. Adams PC, Reboussin DM, Barton JC, McLaren CE, Eckfeldt JH, McLaren GD, Dawkins FW, Acton RT, Harris EL, Gordeuk VR, et al. Hemochromatosis and iron-overload screening in a racially diverse population. N Engl J Med 2005;352(17):1769-78. 28. Næss-Andresen ML, Eggemoen ÅR, Berg JP, Falk RS, Jenum AK. Serum ferritin, soluble transferrin receptor, and total body iron for the detection of iron deficiency in early pregnancy: A multiethnic population-based study with low use of iron supplements. Am J Clin Nutr 2019;109(3):566-575. 29. Barton JC, Wiener HH, Acton RT, Adams PC, Eckfeldt JH, Gordeuk VR, Harris EL, McLaren CE, Harrison H, McLaren GD, et al. Prevalence of iron deficiency in 62,685 women of seven race/ethnicity groups: The HEIRS Study. Connor JR, editor. PLoS One 2020;15:e0232125. 30. Spencer BR, Guo Y, Cable RG, Kiss JE, Busch MP, Page GP, Endres-Dighe SM, Kleinman S, Glynn SA, Mast AE. Iron status and risk factors for iron depletion in a racially/ethnically diverse blood donor population. Transfusion 2019;59:3146– 56. 31. Zacharski LR, Ornstein DL, Woloshin S, Schwartz LM. Association of age, sex, and race with body iron stores in adults: Analysis of NHANES III data. Am Heart J 2000;40(1):98-104. 32. Miller EM. Iron status and reproduction in US women: National health and nutrition examination survey, 1999-2006. PLoS One 2014;9(11):e112216.. 183 33. Godsland IF, Seed M, Simpson R, Broom G, Wynn V. Comparison of haematological indices between women of four ethnic groups and the effect of oral contraceptives. J Clin Pathol 1983;36:184–91. 34. Dekker LH, Nicolaou M, Van Der A DL, Busschers WB, Brewster LM, Snijder MB, Stronks K, Van Valkengoed IGM. Sex differences in the association between serum ferritin and fasting glucose in type 2 diabetes among South Asian Surinamese, African Surinamese, and ethnic Dutch: The population-based SUNSET study. Diabetes Care 2013;36:965–71. 35. Zacharski LR, Shamayeva G, Chow BK, DePalma RG. Racial health disparities, and variant red cell and iron homeostasis. J Health Care Poor Underserved 2016;27:741–61. 36. Pan Y, Jackson RT. Insights into the ethnic differences in serum ferritin between black and white US adult men. Am J Hum Biol 2008;20:406–16. 37. Barton JC, Acton RT, Dawkins FW, Adams PC, Lovato L, Leiendecker-Foster C, McLaren CE, Reboussin DM, Speechley MR, Gordeuk VR, et al. Initial screening transferrin saturation values, serum ferritin concentrations, and HFE genotypes in whites and blacks in the Hemochromatosis and Iron Overload Screening Study. Genet Test 2005;9(3):231-41. 38. Beutler E, West C. Hematologic differences between African-Americans and whites: The roles of iron deficiency and α-thalassemia on hemoglobin levels and mean corpuscular volume. Blood 2005;106:740–5. 39. Perry GS, Byers T, Yip R, Margen S. Iron nutrition does not account for the hemoglobin differences between blacks and whites. J Nutr 1992;122:1417–24. 184 40. Johnson-Spear MA, Yip R. Hemoglobin difference between black and white women with comparable iron status: justification for race-specific anemia criteria1. Am J Clin Nutr 1994;60(1):117-21.. 41. McLaren CE, Li KT, Gordeuk VR, Hasselblad V, McLaren GD. Relationship between transferrin saturation and iron stores in the African American and US Caucasian populations: Analysis of data from the third National Health and Nutrition Examination Survey. Blood 2001;98:2345–51. 42. Meyers LD, Habicht JP, Johnson CL. Components of the difference in hemoglobin concentrations in blood between black and white women in the united states. Am J Epidemiol 1979;109:539–49. 43. Pan WH, Habicht JP. The non-lron-deficiency-related difference in hemoglobin concentration distribution between blacks and whites and between men and women. Am J Epidemiol 1991;134:1410–6. 44. Zacharski LR, Shamayeva G, Chow BK, Depalma RG. Racial differences in iron measures andoutcomes observed during an iron reduction trial in peripheral arterial disease. J Health Care Poor Underserved 2015;26:243–59. 45. Harris EL, McLaren CE, Reboussin DM, Gordeuk VR, Barton JC, Acton RT, McLaren GD, Vogt TM, Snively BM, Leiendecker-Foster C, et al. Serum Ferritin and Transferrin Saturation in Asians and Pacific Islanders. Arch Intern Med 2007;167:722. 46. Ye K, Cao C, Lin X, O’Brien KO, Gu Z. Natural selection on HFE in Asian populations contributes to enhanced non-heme iron absorption. BMC Genet 2015;16:61. 185 47. Barton JC, Acton RT, Lovato L, Speechley MR, McLaren CE, Harris EL, Reboussin DM, Adams PC, Dawkins FW, Gordeuk VR, et al. Initial screening transferrin saturation values, serum ferritin concentrations, and HFE genotypes in Native Americans and whites in the Hemochromatosis and Iron Overload Screening Study. Clin Genet 2006;69:48–57. 48. Pynaert I, Delanghe J, Temmerman M, De Henauw S. Iron Intake in Relation to Diet and Iron Status of Young Adult Women. Ann Nutr Metab 2007;51:172–81. 49. Borch-Iohnsen B, Sandstad B, Åsberg A. Iron status among 3005 women aged 20- 55 years in Central Norway? The Nord-Trøndelag Health Study (the HUNT Study). Scand J Clin Lab Invest 2005;65:45–54. 50. Quintas ME, Requejo AM, Ortega RM, Redondo MR, López-Sobaler AM, Gaspar MJ. The female Spanish population: A group at risk of nutritional iron deficiency. Int J Food Sci Nutr 1997;48:271–9. 51. Milman N, Ulrik CS, Graudal N, Jordal R. Iron status in young Danes. Evaluation by serum ferritin and haemoglobin in a population survey of 634 individuals aged 14-23 yr. Eur J Haematol 1997;58:160–6. 52. Brussaard JH, Brants HAM, Bouman M, Löwik MRH. Iron intake and iron status among adults in the Netherlands. Eur J Clin Nutr 1997;51:51–8. 53. Milman N, Clausen JO, Jordal R. iron status in young Danish men and women: a population survey comprising 548 individuals. Ann Hematol 1995;70(4):215-21. 54. Milman N, Kirchhoff M. Iron stores in 1359, 30- to 60-year-old Danish women: Evaluation by serum ferritin and hemoglobin. Ann Hematol 1992;64:22–7. 186 55. Milman N, Pedersen AN, Ovesen L, Schroll M. Iron status in 358 apparently healthy 80-year-old Danish men and women: Relation to food composition and dietary and supplemental iron intake. Ann Hematol 2004;83:423–9. 56. Lahti-Koski M, Valsta LM, Alfthan G, Tapanainen H, Aro A. Iron status of adults in the capital area of Finland. Eur J Nutr 2003;42:287–92. 57. Milman N, Byg K-E, Ovesen L, Kirchhoff M, Jurgensen KS-L. Iron status in Danish women, 1984-1994: a cohort comparison of changes in iron stores and the prevalence of iron deficiency and iron overload. Eur J Haematol 2003;71:51–61. 58. Fleming DJ, Jacques PF, Tucker KL, Massaro JM, D’Agostino S, Wilson PWF, Wood RJ. Iron status of the free-living, elderly Framingham Heart Study cohort: An iron-replete population with a high prevalence of elevated iron stores. Am J Clin Nutr 2001;73:638–46. 59. Milman N, Byg KE, Ovesen L. Iron status in Danes 1994: II: Prevalence of iron deficiency and iron overload in 1319 Danish women aged 40-70 years. Influence of blood donation, alcohol intake and iron supplementation. Ann Hematol 2000;79:612–21. 60. van de Vijver LPL, Kardinaal AFM, Charzewska J, Rotily M, Charles P, Maggiolini M, Ando S, Väänänen K, Wajszczyk B, Heikkinen J, et al. Calcium Intake Is Weakly but Consistently Negatively Associated with Iron Status in Girls and Women in Six European Countries. J Nutr 1999;129:963–8. 61. Milman N, Ovesen L, Byg K-E, Graudal N. Iron status in Danes updated 1994. I: Prevalence of iron deficiency and iron overload in 1332 men aged 40–70 years. 187 Influence of blood donation, alcohol intake, and iron supplementation. Ann Hematol 1999;78(9):393-400. 62. Galan P, Yoon HC, Preziosi P, Viteri F, Valeix P, Fieux B, Briançon S, Malvy D, Roussel AM, Favier A, et al. Determining factors in the iron status of adult women in the SU.VI.MAX study. Eur J Clin Nutr 1998;52:383–8. 63. Milman N, Taylor CL, Merkel J, Brannon PM. Iron status in pregnant women and women of reproductive age in Europe. Am J Clin Nutr 2017;106(Suppl 6): 1655S– 1662S. 64. Oh SH, Jeong TD, Lee W, Chun S, Min WK. Analysis of Familial Tendencies in Transferrin Saturation in a Korean Population. Dig Dis Sci 2015;60:3136–41. 65. Lee SH, Kim JW, Shin SH, Kang KP, Choi HC, Choi SH, Park KU, Kim HY, Kang W, Jeong SH. HFE gene mutations, serum ferritin level, transferrin saturation, and their clinical correlates in a Korean population. Dig Dis Sci 2009;54:879–86. 66. Wang JL, Shaw NS. Iron status of the Taiwanese elderly: The prevalence of iron deficiency and elevated iron stores. Asia Pac J Clin Nutr 2005;14(3):278-84. 67. Root MM, Hu J, Stephenson LS, Parker RS, Campbell TC. Iron status of middle- aged women in five counties of rural China. Eur J Clin Nutr 1999;53:199–206. 68. Hu PJ, Ley SH, Bhupathiraju SN, Li Y, Wang DD. Associations of dietary, lifestyle, and sociodemographic factors with iron status in Chinese adults: A cross- sectional study in the China Health and Nutrition Survey. Am J Clin Nutr 2017;105(2):503-512. 188 69. He J, Shen X, Fang A, Song J, Li H, Guo M, Li K. Association between predominantly plant-based diets and iron status in Chinese adults: A cross- sectional analysis. Br J Nutr 2016;116:1621–32. 70. Li J, Xiao C, Yang H, Zhou Y, Wang R, Cao Y. Anemia and Iron Status Among Different Body Size Phenotypes in Chinese Adult Population: a Nation-Wide, Health and Nutrition Survey. Biol Trace Elem Res 2018;185:1–10. 71. Shi Z, Hu X, He K, Yuan B, Garg M. Joint association of magnesium and iron intake with anemia among Chinese adults. Nutrition 2008;24:977–84. 72. Tang Q, Liu Z, Tang Y, Tan A, Gao Y, Lu Z, Wang Q, Chen Y, Wu C, Zhang H, et al. High serum ferritin level is an independent risk factor for metabolic syndrome in a Chinese male cohort population. Diabetol Metab Syndr 2015;7:11. 73. Han L ling, Wang Y xia, Li J, Zhang X lei, Bian C, Wang H, Du S, Suo L na. Gender differences in associations of serum ferritin and diabetes, metabolic syndrome, and obesity in the China Health and Nutrition Survey. Mol Nutr Food Res 2014;58:2189–95. 74. Chang JS, Lin SM, Huang T chieh, Chao JCJ, Chen YC, Pan WH, Bai CH. Serum ferritin and risk of the metabolic syndrome: A population-based study. Asia Pac J Clin Nutr 2013;22:400–7. 75. Jung CH, Lee MJ, Hwang JY, Jang JE, Leem J, Park JY, Lee JB, Kim HK, Lee WJ. Elevated Serum Ferritin Level Is Associated with the Incident Type 2 Diabetes in Healthy Korean Men: A 4 Year Longitudinal Study. PLoS One 2013;8(9):e75250. 189 76. Park SK, Ryoo JH, Kim MG, Shin JY. Association of serum ferritin and the development of metabolic syndrome in middle-aged Korean men: A 5-year follow-up study. Diabetes Care 2012;35:2521–6. 77. Kang HT, Linton JA, Shim JY. Serum ferritin level is associated with the prevalence of metabolic syndrome in Korean adults: The 2007-2008 Korean National Health and Nutrition Examination Survey. Clin Chim Acta 2012;413:636–41. 78. Lee BK, Kim Y, Kim Y Il. Association of serum ferritin with metabolic syndrome and diabetes mellitus in the South Korean general population according to the Korean National Health and Nutrition Examination Survey 2008. Metabolism 2011;60:1416–24. 79. Sun L, Franco OH, Hu FB, Cai L, Yu Z, Li H, Ye X, Qi Q, Wang J, Pan A, et al. Ferritin concentrations, metabolic syndrome, and type 2 diabetes in middle-aged and elderly Chinese. J Clin Endocrinol Metab 2008;93:4690–6. 80. Shi Z, Hu X, Yuan B, Pan X, Meyer HE, Holmboe-Ottesen G. Association between serum ferritin, hemoglobin, iron intake, and diabetes in adults in Jiangsu, China. Diabetes Care 2006; 9(8):1878-83. 81. Choi CW, Lee J, Park KH, Yoon SY, Choi IK, Oh SC, Seo JH, Kim BS, Shin SW, Kim YH, et al. Prevalence and characteristics of anemia in the elderly: Cross- sectional study of three urban Korean population samples. Am J Hematol 2004;77:26–30. 190 82. Chang JS, Lin SM, Chao JC-J, Chen YC, Wang CM, Chou NH, Pan WH, Bai CH. Serum ferritin contributes to racial or geographic disparities in metabolic syndrome in Taiwan. Public Health Nutr 2014;17:1498–506. 83. Li X, He T, Yu K, Lu Q, Alkasir R, Guo G, Xue Y. Markers of iron status are associated with risk of hyperuricemia among chinese adults: Nationwide population-based study. Nutrients 2018;10:191. 84. Kim H, Shin C, Baik I. Associations between Lifestyle Factors and Iron Overload in Korean Adults. Clin Nutr Res 2016;5:270. 191