Evidence has emerged in the past few decades that calls into question the need for and the optimal intake levels of a range of essential nutrients. For instance, questions have been raised about the appropriate dose of iron, vitamin D, and calcium in prenatal supplements. There is also a growing recognition that pregnancy is a proinflammatory condition, and inflammation can affect a pregnancy. As diet can influence inflammatory markers, determining optimal antioxidant intake has become an active area of study. Concerns have also been raised about the iodine status of pregnant and lactating women. The third session of the workshop, moderated by Emily Oken, professor in the Department of Population Medicine at the Harvard Medical School and the Harvard Pilgrim Health Care Institute, explored the current state of the science and existing recommendations for these essential nutrients. Highlights from the session presentations are listed in Box 4-1.
Kimberly O’Brien, professor of human nutrition in the Division of Nutritional Sciences at Cornell University, reviewed the state of the science on iron, vitamin D, and calcium intake during pregnancy and lactation. O’Brien touched on current intake recommendations and intake levels, the need for—and outcomes related to—supplementation, and biomarkers related to these nutrients.
Iron Intake Recommendations for Pregnancy and Lactation
O’Brien noted that the Dietary Reference Intakes (DRIs) for iron were established in 2001. The Estimated Average Requirement (EAR) and Recommended Dietary Allowance (RDA) for iron for nonpregnant women are 8.1 and 18 mg/day, respectively. In contrast, the iron EAR and RDA
for pregnant women are 22 and 27 mg/day, respectively. The marked increase in iron intake recommendations during pregnancy is to support red blood cell mass expansion, to provide iron to the placenta, and to transfer approximately 300 mg of iron to the fetus, explained O’Brien. For lactating women, the iron EAR and RDA are lower than the recommendations for nonpregnant women (6.5 and 9 mg/day, respectively). Iron needs decrease during lactation because the iron content of breast milk is relatively low (approximately 300 μg/L) and women who exclusively breastfeed typically experience physiological amenorrhea.
Iron Deficiency During Pregnancy
Different measures of iron status have been used to assess iron deficiency in pregnant women in North America. Using serum ferritin concentrations less than 12 μg/L as the cutoff for iron deficiency, an analysis of the National Health and Nutrition Examination Survey (NHANES) 1999–2010 data found the prevalence of iron deficiency increased from 7 percent during the first trimester to 39 percent during the third trimester (Mei et al., 2011); the prevalence of anemia, however, was markedly lower. O’Brien showed that the Alberta Pregnancy Outcomes and Nutrition (APrON) Study, a longitudinal cohort in Canada, had similar findings—40 percent of women in their third trimester were considered iron deficient, but only 5 percent were classified as anemic. An NHANES 1999–2010 analysis found that total body iron stores are depleted over the course of gestation, with nearly 30 percent of women in their third trimester having completely depleted their iron stores (Gupta et al., 2017).
O’Brien described how iron homeostasis is maintained during pregnancy. Evidence from a stable isotope study suggests that absorption of non-heme iron increases as serum ferritin concentrations decrease (O’Brien et al., 1999), but she noted that relative increases in iron absorption are modest, as our bodies do not have a physiologically regulated pathway to excrete excess iron. New evidence in the past few decades has revealed that the hormones hepcidin and erythroferrone, in addition to erythropoietin, are involved in the regulation of maternal iron homeostasis. Increased attention has also been placed on the placenta and its ability to respond to maternal and fetal iron status. A study found that the placenta can respond to low maternal iron status by upregulating iron uptake proteins (Young et al., 2010). A recent study suggests, however, the placenta may prioritize iron for its own use before providing iron to the fetus (Sangkhae et al., 2020). Evidence on adaptations to maintain iron homeostasis that occur during pregnancy raise questions about the appropriate dose of iron in prenatal supplements, suggested O’Brien.
Prenatal Iron Supplementation
Recent reviews have assessed the possible risks and benefits of prenatal iron supplementation. A U.S. Preventive Services Task Force report (McDonagh et al., 2015) and Cochrane reviews (Peña-Rosas et al., 2012, 2015) all concluded that routine prenatal iron supplementation improves maternal hematological indices, but the clinical significance and effects on maternal and infant outcomes are unclear. O’Brien stated that one challenge of translating evidence into policy and practice is that studies often use a form of iron (ferrous sulfate) that is different from the form available in prenatal supplements (fumarate or gluconate) (Saldanha et al., 2019b).
The iron content of prenatal supplements in the United States and Canada slightly differ from each other, noted O’Brien. In the United States, prenatal supplements contain 27 mg of iron, whereas Canadian prenatal supplements contain 16–20 mg of iron. O’Brien explained that the lower iron content of the Canadian supplements stems from an analysis of data from the Canadian Community Health Survey (Cockell et al., 2009). In the analysis, the distribution of usual iron intake was determined using dietary recall data from nonpregnant women. Average iron intake of nonpregnant women fell well below the EAR for pregnant women. It was estimated that an iron supplement of 16 mg/day would shift the intake distribution curve such that less than 3 percent of women would have an iron intake falling below the EAR. A 20 mg/day iron supplement would shift women above the EAR, and less than 1 percent of the sample would have an intake above the Tolerable Upper Intake Level (UL). An iron supplement of 27 mg/day, however, was estimated to shift the usual intake distribution curve such that one-third of women would have intakes in excess of the UL.
Suggesting that the analysis of the Canadian Community Health Survey analysis treated iron as a single pool, O’Brien cautioned that the body treats heme and non-heme iron differently and that both quantity and composition of iron intake need to be considered. The vast majority of iron consumed in the diet is non-heme iron, said O’Brien. Previous estimates evaluated dietary needs assuming that only 25 percent of non-heme iron is absorbed. These estimates do not account for the much higher absorption of heme iron. In pregnant and nonpregnant women, approximately 50 percent of dietary heme iron is absorbed. Unlike non-heme iron, absorption of heme iron was not downregulated in iron-replete women (Young et al., 2010). O’Brien explained that the composition of dietary iron intake varies by dietary patterns. For instance, a woman following a ketogenic diet will consume different proportions and quantities of heme and non-heme iron compared to a woman consuming a plant-based diet. These dietary differences can have implications for interpreting the results of iron supplementation trials, indicated O’Brien. She further noted that iron supplementation
trials often do not account for the presupplementation iron status of participants, which can also affect results.
Iron-Deficiency Anemia During Pregnancy
Pregnant women who have anemia are often advised to ingest an additional 40–200 mg/day of supplemental iron, said O’Brien. Absorption of non-heme iron decreases with increasing dose. When iron is given as a large dose, much of it will remain unabsorbed and continue through the intestinal tract. O’Brien suggested that the unabsorbed iron likely has implications for the gut microbiome and acknowledged that more studies are needed to investigate this relationship.
Hemoglobin thresholds for anemia during pregnancy currently in use were established in 1989 by the Centers for Disease Control and Prevention (CDC). Lower thresholds are used during the second trimester owing to hemodilution, and lower thresholds are recommended for black adults throughout pregnancy. O’Brien explained that the trimester-specific CDC thresholds for anemia were established based on the fifth percentile of hemoglobin at 12, 20, and 32 weeks gestation. These hemoglobin distributions were established using data from four studies conducted in Europe in less than 400 women. O’Brien asserted that participants in those studies differ from the current composition of the North American population. For instance, two of the studies reported average body mass indexes of 18 and 21 kg/m2, which is markedly different from what currently exists in the population. O’Brien noted that an excess of adipose may increase systemic inflammation, which in turn, may increase hepcidin production. She suggested that prolonged elevated hepcidin exposure may decrease iron absorption. The studies used to establish the CDC anemia thresholds included primarily Caucasian participants who had iron intakes above current recommendations, which O’Brien underscored as further evidence that the data do not necessarily reflect the current circumstances.
Only a small portion of anemia cases are classified as iron-deficiency anemia. O’Brien indicated that this may be attributable to inaccurate hemoglobin thresholds, inaccurate cutoffs for iron status biomarkers, or inappropriate adjustments for stage of gestation, expansion of plasma volume, or inflammation. Deficiencies in other nutrients may also affect anemia risk. A study of pregnant adolescents found those with low 25-hydroxy vitamin D concentrations had eight times the risk of anemia at term, compared to those with higher vitamin D status.
Certain population groups are at higher risk of iron deficiency during pregnancy. O’Brien noted that iron deficiency is more common later in pregnancy, among women carrying multiples, among non-Hispanic blacks, and among pregnant adolescents. She reiterated that women with pre-
existing inflammation may also be at risk for compromised iron status owing to the effects on hepcidin but stated that more research is needed. Some population groups are also at risk for excessive iron status; the highest prevalence was found among Asians and Pacific Islanders.
Research Needs Related to Iron
Despite evidence that iron supplementation improves hematological indicators, more research into the benefits and risks of supplementation is warranted, O’Brien said. In particular, she said that more work is needed to determine an optimal dose of iron supplementation during pregnancy, to evaluate current thresholds for hemoglobin and biomarker data that determine iron status, and to identify characteristics of at-risk populations (e.g., racial and ethnic differences) in order to improve screening.
Vitamin D Needs, Status, and Intake During Pregnancy and Lactation
Moving on to vitamin D, O’Brien explained the DRIs for vitamin D were last updated in 2011. The EARs and the RDAs for adults were based on evidence related to bone health. No additional vitamin D was recommended for pregnant women because evidence did not support an association between vitamin D status and bone mass in pregnancy, and animal studies did not provide evidence related to vitamin D intake and fetal development. Vitamin D intake recommendations were also not increased for lactating women because vitamin D status did not appear to be related to bone mineral density or the calcium or vitamin D composition of breast milk. A wide variety of maternal and fetal outcomes have been investigated in relation to vitamin D beyond bone health. “When you are talking about how much vitamin D do we require, it is very important to think about the outcome,” said O’Brien.
Vitamin D status has been assessed in both the U.S. and Canadian populations. Analysis of the NHANES 2009–2010 data found that among females 12 years of age and older 8 percent had 25-hydroxyvitamin D concentrations less than 12 ng/mL (indicating risk of deficiency) and 28 percent had concentrations less than 20 ng/mL (indicating risk of deficiency or inadequacy) (Schleicher et al., 2016). Similar estimates were found among pregnant and lactating women in the 2001–2006 NHANES cycles (Looker et al., 2011). Non-Hispanic blacks were found to be at the highest risk for vitamin D inadequacy or deficiency (Herrick et al., 2019). Estimates of vitamin D inadequacy among females in Canada were similar to those found in the United States; the 2016–2017 cycle of the Canadian Health Measures Survey reported 24 percent of females 3–79 years of age had 25-hydroxyvitamin D concentrations less than 20 ng/mL, said O’Brien.
Vitamin D Biomarkers During Pregnancy
The biomarker 25-hydroxyvitamin D is typically used to assess vitamin D status; however, there are unique changes in vitamin D biomarkers that are evident only during pregnancy. During pregnancy, the hormone 1,25-dihydroxyvitamin D and the concentrations of vitamin D binding protein increase. Both 1,25-dihydroxyvitamin D and 24,25-dihydroxyvitamin D are influenced by 25-hydroxyvitamin D status. The placenta also expresses enzymes capable of activating and inactivating 25-hydroxyvitamin D.
The evidence is less clear on the effects of these changes to vitamin D metabolism during pregnancy. O’Brien indicated that there does not appear to be a link between elevated 1,25-dihydroxyvitamin D and calcium absorption and thought there are likely immunomodulatory roles of this hormone that need further characterization. She emphasized the importance of identifying systemic and cellular-level effects of these vitamin D metabolites during pregnancy to inform intake recommendations. “If [25-hydroxyvitamin D] is the only biomarker of interest when survey studies are done, you really are missing out on what is happening with partitioning of this prohormone into these products,” said O’Brien.
Vitamin D Supplementation During Pregnancy
At present, there is no consensus as to whether vitamin D should be supplemented during pregnancy, O’Brien said. Vitamin D trials during pregnancy differ from each other with respect to when in pregnancy supplementation begins, what dose of vitamin D supplement is used, and whether the supplement contains other nutrients. Certain key variables, such as prepregnancy body mass index, are not always accounted for and may affect interpretation of the results. O’Brien said that these inconsistencies make it difficult to compare studies.
Calcium During Pregnancy and Lactation
Like vitamin D, the calcium DRIs were updated in 2011; the updated DRIs were informed by evidence related to bone health and do not recommend an increase in needs during pregnancy or lactation, explained O’Brien. Calcium absorption increases during pregnancy to meet fetal calcium demands. Changes in calcium metabolism that occur during pregnancy and lactation do not appear to have a negative effect on long-term bone health. Bone mineral density does not appear to decrease in women with increased parity or in those with extended breastfeeding, and bone loss from lactation appears to be recovered.
Many of the calcium supplementation trials conducted in pregnant women have focused on the outcome of preeclampsia, rather than measures of maternal or infant bone health, noted O’Brien. She indicated that more data are needed to evaluate the effect of calcium supplementation on bone health in pregnant women at the extremes of the reproductive age groups and that there may be other population groups at risk, but she characterized the data on bone mass during pregnancy as limited.
O’Brien offered ideas for future research directions related to calcium. With techniques that can assess fetal bone development and maternal bone status, she suggested that there is the potential to better understand bone morphology and growth over the course of gestation. O’Brien also put forth the idea that new approaches to assessing maternal, fetal, and placental responses to nutrient exposures were needed. Finally, she thought the “omic” and multi-omic technologies might help identify responses to nutrients not yet characterized.
Corrine Hanson, director and professor in the Medical Nutrition Education Division at the University of Nebraska Medical Center, acknowledged that the term antioxidants encompasses hundreds of compounds. Hanson, therefore, focused her remarks on the current state of the science on the relationship between women’s intake and status of select dietary antioxidants and pregnancy and infant outcomes.
Overview of Oxidative Stress
To set the stage for her remarks, Hanson explained that reactive species produce free radicals that can create DNA and cellular damage. She defined oxidative stress as “an imbalance between the production of various reactive species and the ability of the organisms’ natural protective mechanisms to cope with these reactive compounds and prevent adverse events” (IOM, 2000). According to Hanson, the body has various protective mechanisms against oxidative stress, with diet as one of the best defenses.
A dietary antioxidant is as a substance found in the human diet in commonly consumed foods that significantly counteracts the adverse effects of reactive oxygen and/or nitrogen species in vivo in humans (IOM, 2000). Dietary antioxidants that meet these formal criteria include vitamin C, vitamin E, and selenium. Hanson noted that vitamin E has eight isoforms, with
alpha-tocopherol being the form that is considered to meet dietary vitamin E requirements in humans. Other compounds—including beta-carotene, alpha-carotene, beta-cryptoxanthin, lutein, lycopene, and zeaxanthin—do not technically meet the formal criteria, but they are typically considered dietary antioxidants given their influence on reactions related to the oxidative process. Polyphenols also do not meet the criteria, but they are also treated as dietary antioxidants, added Hanson.
Antioxidant Intake Recommendations and Relationships to Status
DRI values for adequacy have been established for vitamin C, vitamin E, and selenium based on their antioxidant properties. The RDAs for vitamin C are the same for pregnant and lactating females (80 mg/day for adolescents and 85 mg/day for adult women). In contrast, the RDAs for vitamin E for pregnant women are lower than for lactating women (15 versus 19 mg/day, respectively), as is the case for selenium (60 versus 70 μg/day, respectively). For carotenoids, the available evidence at the time was considered insufficient to establish DRIs for adequacy. Hanson remarked that the lack of a recommendation for adequacy makes it difficult to assess carotenoid intake. She also emphasized the importance of understanding antioxidant requirements during pregnancy, because newborns, particularly high-risk premature infants, encounter a range of exposures that can induce oxidative stress.
Relationships between intakes of some of the dietary antioxidants and maternal and infant serum concentrations have been assessed. Infant serum carotenoid concentration is correlated with maternal intake, but it is only 5–15 percent of maternal serum concentration. Hanson emphasized that it is difficult to determine if the carotenoid values are truly low without a quantitative reference based on carotenoids’ antioxidant properties. Maternal intake and infant status is correlated for some of the isoforms of vitamin E (e.g., gamma-tocopherol) but not for alpha-tocopherol.
Hanson explained that different forms of a nutrient may have opposite effects. She drew on two isoforms of vitamin E as an example. Alpha-tocopherol—found in leafy greens, nuts, and seeds—is classified as a dietary antioxidant. In contrast, gamma-tocopherol, found in all processed foods and fast foods that contain soy oil, is being investigated for its pro-oxidant and proinflammatory properties. “At this time, we have not defined any nutrition compounds as pro-oxidants,” said Hanson, but noted that this is an active area of research.
Predictors of Antioxidant Status During Pregnancy
Describing Nebraska as a rural, agricultural state, Hanson indicated that health disparities exists that are putting certain population groups at
risk for poor maternal–child outcomes. To better understand predictors of antioxidant status during pregnancy, Hanson’s group studied a cohort of 500 pregnant women in Nebraska. Her team found that serum concentrations of lycopene, beta-carotene, and beta-cryptoxanthin are lower among the women who lived in a food desert, and noted that 20 percent of the cohort lived in a food desert. The study also revealed serum concentrations of lutein, beta-carotene, and alpha-tocopherol were higher, and concentrations of gamma-tocopherol lower, among women with private insurance, as compared to those who use public insurance. Hanson reported that 60 percent of her cohort used public insurance. Nonwhites were 3.5 times more likely to be vitamin E deficient as compared to their white counterparts, and women with some degree of food insecurity were 5.5 times more likely to have vitamin E intakes below the RDA.
Outcomes Related to Antioxidant Status and Intake During Pregnancy
To contextualize the importance of the disparities her group found in its cohort study, Hanson provided an overview of evidence relating antioxidant status and intake to premature delivery, preeclampsia, respiratory issues, and growth.
Both observational and trials have explored the relationship between antioxidants and premature delivery. Observational studies have reported that lower plasma levels of carotenoids and antioxidants and higher levels of gamma-tocopherols are associated with increased risk of prematurity (Carmichael et al., 2013; Kramer et al., 2009). Vitamin C intakes below the 10th percentile were found to double the risk of preterm delivery (Siega-Riz et al., 2002). A supplementation trial in Turkey found that women who received a supplement containing vitamins C and E (1,000 and 400 mg/day, respectively) stayed pregnant longer when they experienced premature rupture of membranes, as compared to women in the control group (Gungorduk et al., 2014). Large trials of antioxidant supplementation conducted in the United States and in the United Kingdom, however, did not find an effect on preterm birth. Hanson raised the issue of participant antioxidant status for these trials, as serum concentrations were either not reported or indicated that the women were replete.
Moving on to evidence on relationships with preeclampsia, Hanson reported that findings from observational studies are inconsistent and supplementation trials did not find an effect. One secondary analysis of a trial, however, found preeclampsia rates were lower among smokers who received the supplement (Abramovici et al., 2015). One study found that women in the lowest tertile of selenium status had a decrease in the preeclampsia biomarker sFlt-1 when given a selenium supplement (Rayman et al., 2014). A trial is currently under way further exploring this finding, Hanson said.
Studies have also looked at antioxidant intakes and childhood respiratory outcomes. A meta-analysis of studies conducted in developed countries reported that vitamin E intake during pregnancy appears to have a protective effect on respiratory outcomes in children, primarily childhood wheeze, but most studies did not find such a relationship with vitamin C intake (Beckhaus et al., 2015). Hanson cautioned that these findings were based on intake data, rather than serum measurements. Secondary analyses of two large supplementation trials did not find a relationship with childhood respiratory outcomes, said Hanson.
Infant growth has been investigated as an outcome of interest related to antioxidant status and intake. One observational study found higher birth weight percentile, head circumference, and length percentile among children born by women with higher total serum lycopene concentrations. Hanson remarked that in her group’s cohort, higher serum concentrations of any isoform of vitamin E (including gamma-tocopherol) was associated with higher weight, length, and head circumference at birth. However, a trial reported that risk of intrauterine growth restriction was not reduced with prenatal vitamin C and vitamin E supplementation (Rumbold et al., 2006).
Despite the potential for benefits, higher antioxidant exposure could potentially lead to harm. A case–control study reported that vitamin E intake was higher among mothers whose infants were born with a congenital heart defect compared to controls (13.3 versus 12.6 mg/day, respectively; Smedts et al., 2009). Congenital heart defects have also been reported to be nine times higher among women who have high dietary vitamin E intake (> 14.9 mg/day) and used vitamin E supplements periconceptionally, said Hanson. She noted that these findings were from observational studies.
“There have been huge shifts in our iodine intake in the last few decades,” said Elizabeth Pearce, professor of medicine at the Boston University School of Medicine in the Section of Endocrinology, Diabetes, and Nutrition. To review the importance of the changing landscape of iodine nutrition and its relevance to pregnancy and lactation, Pearce discussed dietary iodine requirements and status assessment, the consequences of iodine deficiency or excess, current U.S. iodine nutrition status, sources of iodine in the diet, and current recommendations for iodine-containing supplements.
Dietary Iodine Requirements and Status Assessment
The only known use of iodine in the body is for thyroid hormone synthesis. Thyroid hormone production increases approximately 50 percent during pregnancy, which requires an extra 50–100 μg of iodine as a
substrate. The fetal thyroid also needs iodine to support thyroid hormone synthesis when it begins working in the second half of pregnancy. Maternal iodine readily crosses the placenta and is transferred to the fetus, noted Pearce. These increased needs during pregnancy are coupled with increased losses of iodine through the urine. Pregnant women lose 30–50 percent more iodine than nonpregnant adults. For lactating women, iodine is actively secreted into breast milk.
Iodine intake recommendations have been published by the Institute of Medicine (IOM, 2006) and a joint effort from the World Health Organization (WHO), the United Nations Children’s Fund (UNICEF), and the International Council for Control of Iodine Deficiency Disorders (ICCIDD) (WHO/UNICEF/ICCIDD, 2007). Both publications recommend iodine intake of 150 μg/day for nonpregnant adults, whereas the recommendations for pregnancy range from 220–250 μg/day and from 250–290 μg/day for lactating women. Pearce indicated that the higher iodine intake recommendations for lactating women have recently been called into question.
Median urinary iodine excretion can be used to characterize population-level iodine status. For nonpregnant adults, a median urinary iodine concentration of 100–199 μg/L is considered optimal (WHO Secretariat et al., 2007). The optimal median urinary iodine concentration range is 150–249 μg/L during pregnancy and at least 100 μg/L during lactation (WHO Secretariat et al., 2007). Urinary iodine excretion cannot be used to assess individual status, said Pearce. She explained that iodine intake tends to be episodic, leading to highly variable excretions in the urine (Als et al., 2000). “It has been estimated that you really need 10 to 12 ideally 24-hour urine collections to understand chronic intakes with even 20 percent precision. I have yet to meet a patient who would actually do this,” remarked Pearce.
Consequences of Deficient Iodine
Inadequate iodine intake at any life stage can have detrimental effects, including hypothyroidism and an enlargement of the thyroid, called goiter. Pregnancy, however, is associated with the most adverse effects caused by iodine deficiency, noted Pearce. Severe iodine deficiency during pregnancy can lead to cretinism, a syndrome in which the child’s growth and intellectual development is impaired. Given the critical role the thyroid hormone plays in fetal brain development, even modest iodine deficiency during pregnancy is thought to effect children’s neurocognition. Severe maternal deficiency has been associated with increased risk for negative obstetric outcomes, including miscarriage, stillbirth, and perinatal and infant mortality.
Relationships between maternal iodine intake during pregnancy and children’s neurodevelopment have been assessed. For instance, some studies
have reported that low maternal iodine intake may increase a child’s risk for attention deficit hyperactive disorder and may possibly be linked to autism spectrum disorders. Several studies have assessed the effect of maternal iodine nutrition on intelligence. Animal model data suggest that the first trimester of pregnancy is the most critical window for maternal iodine nutrition. However, iodine deficiency at any point when the brain is rapidly developing—from gestation through the first 3 years of life—can lead to intellectual impairments, indicated Pearce. A meta-analysis found that children of women with optimal iodine nutrition entering pregnancy have intelligence quotient (IQ) benefits of 7.4 points (Bougma et al., 2013). IQ benefits as high as 12–13 points were reported in a study conducted in China that optimized maternal iodine status in populations with severe iodine deficiency. A cohort study in the United Kingdom found a linear dose–response relationship between maternal iodine excretion in the first trimester and children’s IQ at 8 years of age, including among those with mild to moderate iodine deficiency (Bath et al., 2013). “This was a bit concerning because … pregnant women in the United States are now in this mild to moderate deficiency range and that is true actually for much of Western Europe,” Pearce said to contextualize the study findings. She went on to note that other cohort studies have reported similar findings, particularly related to language development (Abel et al., 2017a,b; Hynes et al., 2013; Markhus et al., 2018).
Questions remain as to whether maternal iodine supplementation can improve children’s intellectual outcomes, acknowledged Pearce. A systematic review reported that controlled trials evaluating such outcomes are lacking, particularly among those with mild to moderate iodine deficiency (Taylor et al., 2014). Pearce explained one of the challenges to conducting such a study is that a placebo arm has been considered unethical in some regions. A recent randomized trial conducted in Bangalore, India, and Bangkok, Thailand, sought to provide clarity on the effects of iodine supplementation among mildly deficient pregnant women and IQ. Women who were approximately 10 weeks gestation were randomized to receive a 200 μg/day iodine supplement or a placebo. The trial found null results on children’s IQ at 5 years of age (Gowachirapant et al., 2017). One of the issues the trial encountered was that in the period between planning and implementing the study, India’s salt iodization programs expanded to the extent that the study participants from India were iodine replete at baseline.
Consequences of Excessive Iodine
There is some evidence that iodine has a U-shaped relationship with thyroid function, said Pearce. Most people have homeostatic mechanisms called the Wolff-Chaikoff effect to protect them against large variations in iodine
intake. However, this mechanism can fail in some people, leading to hypothyroidism after a large dose of iodine. She went on to explain that the fetus does not develop the iodine homeostatic control until 36 weeks gestation, so a mother may be able to maintain normal thyroid function for herself after a large dose of iodine, but a fetus could selectively develop hypothyroidism.
There is a lack of population-level data to support how much iodine can be safely consumed. Pearce indicated a study from China offers the best insight (Shi et al., 2015). The study assessed the urinary iodine concentrations of more than 7,000 women in their first trimester and risk of hypothyroidism. The investigators found a U-shaped relationship between urinary iodine excretion and thyroid function, with the lowest risk of hypothyroidism among women with urinary iodine excretion 150–250 μg/L. Pearce suggested this finding could have implications for downstream effects on a child’s development.
Countries have different upper levels of safety for iodine intake. In the United States, the UL for all adults, including pregnant and lactating women, is 1,100 μg/day (IOM, 2006). By contrast, a WHO, UNICEF, and ICCIDD report recommends a limit of 500 μg/day for pregnant and lactating women (WHO/UNICEF/ICCIDD, 2007).
Iodine Nutrition Status in the United States
A century ago, there was a large portion of the country known as “the goiter belt.” The majority of children in the region had palpable goiters. Pearce explained that during World War II, a substantial number of men from Michigan were disqualified from service because of visible goiter or intellectual impairment attributable to maternal iodine deficiency. Issues like these resolved with the introduction of salt iodization in the 1920s, she noted.
The first NHANES, conducted in the 1970s, found that median urinary concentrations of iodine exceeded 300 μg/L, a level considered excessive by WHO. By the third NHANES in the late 1980s and early 1990s, urinary iodine concentrations were halved, for reasons still unknown. There are subgroup differences in urinary iodine excretion by age and sex. The group most at risk for low urinary iodine excretion are women of reproductive age, who are the population group most vulnerable to the effects of iodine deficiency, observed Pearce. A recent NHANES analysis has found that pregnant women in the United States are considered mildly iodine deficient (Perrine et al., 2019).
Sources of Iodine in the Diet
Pearce admitted that it is difficult to determine sources of iodine in the diet. According to 2008–2012 data from the U.S. Food and Drug Admin-
istration Total Diet Study, dairy foods are the largest contributor to iodine intake, followed to a lesser extent by grains. Pearce pointed out that the Total Diet Study does not account for use of iodized salt. She continued, noting that nearly three-quarters of the salt consumed in the United States comes from commercial or processed foods in which the salt is usually not iodized, as salt iodization is voluntary in the United States. The only iodine people consume from salt is that which is added at home or at the table, although most kosher salt and sea salts do not contain iodine. Women of reproductive age are the group least likely to add salt to their food, limiting their iodine intake.
Using bread and milk as examples, Pearce described how the iodine of the food supply can change. With respect to bread, iodate dough conditioners were used to improve bread structure beginning in the 1940s. By the 1970s their use markedly decreased, although some products still include them today. Turning to milk, Pearce explained that on average a glass of milk provides 110 μg of iodine. In the 1940s, farmers started to add iodine to cattle feed to improve their cow milk supplies. Concerns over excessive iodine intakes grew in the 1980s, leading to federal limits on the iodine content of cattle feed. Pearce suggested that this change could possibly explain why urinary iodine excretions were elevated among participants in the first NHANES.
Americans who do not consume dairy are at risk of iodine deficiency, noted Pearce. A study conducted in Boston found that urinary iodine excretion among vegetarians was slightly below the median U.S. excretion level, but vegan participants were markedly lower (Leung et al., 2011).
Current Recommendations for Iodine-Containing Supplements
Since 2006, the American Thyroid Association has recommended that women who are planning to become pregnant, who are pregnant, or who are lactating should take a supplement containing 150 μg/day of iodine. Pearce stated that similar recommendations have since been released from national and international professional groups. The recommendations, however, are not necessarily implemented. A recent NHANES analysis found that less than 20 percent of pregnant women use an iodine-containing supplement (Gupta et al., 2018). Pearce went on to say that only 60 percent of the U.S. multivitamin market share and only a small portion of prescription vitamins contain iodine (Lee et al., 2017).
Hanson, O’Brien, and Pearce took part in a session discussion, moderated by Oken. Questions were raised about the interface between dietary
intake and supplements, considerations related to antioxidants, and implications of low and high iodine intake.
Interface Between Dietary Intake and Supplements
To open the discussion, Oken noted that shifts in dietary patterns could potentially jeopardize nutrient intake and asked each of the speakers what they would recommend to address the changing landscape of shortfall intakes. From her perspective, Pearce endorsed universal salt iodization to optimize iodine status. She acknowledged that public health initiatives recommend reductions in salt intake to reduce cardiovascular disease risk, but she suggested that salt will never be absent from the diet. O’Brien thought that pregnant and lactating women, whose diets are inherently more restrictive, could find ways to incorporate potential shortfall nutrients with the help of a dietitian. Reiterating a point made in her presentation, Hanson explained that for many of the antioxidant compounds, it is difficult to know what intakes are optimal because they do not have quantitative intake recommendations.
Several of the webinar audience members asked whether intake of nutrients from foods should be prioritized over intakes from supplements. Hanson, O’Brien, and Pearce agreed that, ideally, intakes should come from food sources. O’Brien added that supplements should be considered when dietary intakes do not meet nutrient needs. Given nutrient interactions, however, she cautioned against selectively focusing on a single nutrient as multiple micronutrient deficiencies often coexist and high intakes of one nutrient may interfere with absorption of other key nutrients. With respect to iodine, Pearce noted that it is difficult to know where it is coming from in the diet, because it is not on the food label.
Building on the topic of supplementation, Anna Maria Siega-Riz of the University of Massachusetts Amherst asked for thoughts on companies that offer personally designed prenatal vitamins. Hanson thought that the companies benefit financially from telling clients that they need a supplement and suggested that inadequate intakes could largely be corrected by dietary changes. She also raised concerns about the potential for negative outcomes related to supplementation. Hanson said:
Nutrients are not benign. These are pharmacologically active compounds that have very powerful consequences in the body. Do not assume that more is better, and do not assume that when you ingest them, they do not do anything. Be cautious. Treat it as you would a medication.
Related to the concept of personalized prenatal vitamins, Oken wondered if there would be a movement toward creating supplements optimized
for different stages of gestation. O’Brien suggested tailoring in such a way would be challenging and there would be a potential for making changes with implications not yet known to us. Regarding supplement composition, Johanna Dwyer of the National Institutes of Health’s (NIH’s) Office of Dietary Supplements remarked from the audience that the government has compiled a database with the labels of more than 100,000 dietary supplements available on the market in the United States. She also noted that the U.S. Department of Agriculture’s Agricultural Research Service, NIH, and other federal partners have compiled a dietary supplement ingredient database.
Considerations Related to Antioxidants
With respect to dietary antioxidants, Dwyer asked how nonnutrient bioactives, such as butylated hydroxyanisole (BHA) and butylated hydroxytoluene (BHT) should be considered. Hanson admitted that there are a wide range of compounds with antioxidant properties, many of which lack analytical techniques and standards needed to assess whether intakes are high or low.
A member of the webcast viewing audience wanted to know if women who are trying to conceive should take antioxidant supplements. Hanson said that there are several reasons why antioxidant supplementation is not recommended. Many antioxidants do not have DRI values, which limits the ability to determine intake targets, including for needs and for safety, she explained. Given evidence of U-shaped relationships for other nutrients that were discussed, Hanson cautioned that more is not always better. For instance, tocopherol levels are relative to each other, such that increasing one will decrease another; the consequences of such changes, however, are not known. Hanson stated, “There is just so much we do not know that, at this point, supplementation could potentially do more harm than good, and we really recommend that people try to get this from food sources.” Related to this concept, a webcast viewer wondered if pregnant women who are depleted should take antioxidant supplements. Hanson said that antioxidant deficiency is not well characterized, except for the case of selenium for which there is evidence from developing countries where the selenium content of the soil is deficient.
Implications of Low and High Iodine Intake
In reflecting on Pearce’s presentation, Dwyer wanted to know if there are data on the iodine status of Asian women, particularly those who migrated from areas in which the population has low iodine status. Describing the U.S. data on subgroups as limited, Pearce stated the only difference
that could be identified was for African American women, who appear to have lower iodine statuses than other groups.
Regarding iodine intakes and recommendations, a member of the webcast viewing audience asked why the RDA for iodine was significantly lower than the intakes found in the first NHANES and if there were any negative ramifications from such high intakes. Pearce indicated that an entire generation or two of children were exposed to high iodine intakes. Data on the effects of such intakes are limited, noted Pearce, but she indicated that the prevalence of goiter in the United States did not change during this time.