The Impact of Perchlorate Exposure in Early Pregnancy: Is It Safe to Drink the Water?

A large number of environmental agents have been identified that interfere with thyroid function, including those that target thyroid hormone production, metabolism, transport, and action (1,2). Perchlorate (ClO₄⁻) is an inorganic anion that blocks iodine uptake into the thyroid and, at sufficiently high levels and duration of exposure, impairs thyroid hormone production (3). It has been used therapeutically in limited clinical situations, such as for hyperthyroidism associated with amiodarone (typical dose up to 1000 mg/d along with antithyroid drug for 40 d) (3,4). Perchlorate is water soluble and very stable and is found in municipal water supplies, in crops irrigated with perchlorate-containing water, and even in breast milk (3,4). It is used in manufacturing solid rocket fuel, matches, airbag inflation systems, and fireworks. There is also natural formation of perchlorate in the atmosphere that accumulates in groundwater and soil, and it has been detected around the world (3,4). The principal areas of concern over the health consequences of perchlorate exposure have been the effects in susceptible populations such as pregnant women and young children, the role of low iodine intake in increasing perchlorate toxicity, and the permissible levels of perchlorate in the water supply (2,3,4,5).

Perchlorate has a well-characterized site of action, which is direct inhibition of the sodium iodide symporter (NIS), the membrane protein that concentrates iodine in the thyroid gland and other iodine-concentrating tissues, such as the lactating breast (6). Perchlorate competes with iodine and is itself transported into the thyroid by NIS (6). There are other inhibitors of NIS—thiocyanate, contained in cigarette smoke, and nitrates (7). These other NIS inhibitors are also found widely in the environment and are additive to each other in blocking iodine uptake, although perchlorate is the most potent (7).

In the adult, the thyroid has a number of mechanisms to compensate for interference with function and can preserve normal thyroid production across a range of dietary iodine intake (8). In iodine-sufficient regions, most patients who undergo surgical lobectomy maintain normal thyroid function (9), although the literature in this area is controversial. Pregnancy, however, is associated with increased iodine requirements, and even mild disruption of thyroid function has been associated with adverse outcomes in pregnancy and in neurodevelopmental measures in offspring (8,10).

The first measurable effect of increasing doses of perchlorate on the thyroid is reduction in iodine uptake (3,4,5). Short-term studies of the health implications of perchlorate exposure, both in normal volunteers and those with industrial exposure, have shown that perchlorate doses of 0.01 to 1 mg/kg · d reduce iodine uptake approximately 10–80% (3,4,5). Perchlorate exposure below this dose, a “no-effect” level on iodine uptake, was determined to be a dose of 0.007 mg/kg · d (equivalent to about 240 μg/liter perchlorate in drinking water, if water is the only source) (5). These studies, however, have been short-term and in healthy adults, the group most likely to successfully compensate for agents that interfere with thyroid function (2,9). An epidemiological longitudinal study compared pregnant women from cities in northern Chile, one with high perchlorate exposure (average urine perchlorate on first prenatal visit, 132.9 μg/liter), to pregnant women in two cities with lower perchlorate exposure (average urinary perchlorate on first prenatal visit, 24.5 and 66.7 μg/liter) (11). The women were followed during pregnancy for thyroid function, as well as for thyroid function tests and measures of growth of their offspring. Pregnant women exposed to high levels of perchlorate, compared with women with lower perchlorate exposure, did not have any differences in their thyroid function or that of their offspring. The conclusion of the study was that fairly high levels of perchlorate exposure did not influence thyroid function, even in pregnancy. The pregnant women, however, had relatively high iodine intake (average urine iodine, 269 μg/liter), compared with that in the United States (women of reproductive age, median urine iodine, 139 μg/liter) (12), raising the concern that high iodine intake may have protected the women from the consequences of perchlorate. Subsequent evaluation of the data, however, addressing the influence of perchlorate on thyroid function in individuals based on their iodine intake, showed no association (13).

The findings in the pregnancy study from Chile were reassuring regarding the health impact of perchlorate exposure but contrasted with the results of a study from a population in the United States. This study, by Blount et al. (14), used data from the National Health and Nutrition Examination Survey (NHANES 2001–2002) and found that in women who were iodine sufficient (urinary iodine >100 μg/liter), urinary perchlorate levels were positive predictors of serum TSH level. In women with low iodine intake (urinary iodine <100 μg/liter), urinary perchlorate levels were a positive predictor of serum TSH and a negative predictor of serum T₄, although TSH and T₄ levels remained in the reference range. In men, there was no association of perchlorate exposure with thyroid function. The influence of iodine intake on susceptibility to perchlorate effects on thyroid function from this study was compelling and raised concerns about both sufficient iodine intake in women of reproductive age and the risk of perchlorate exposure.

The essential unanswered question about the health consequences of perchlorate has been the level of perchlorate likely to have an effect on pregnant women with low iodine intake, which is seen in as many as 9% of pregnant women in the United States (2,3,4,5). The study by Pearce et al. in this issue (15) directly addresses this question. The cross-sectional study used samples from the Controlled Antenatal Thyroid Screening Study (CATS), which is being conducted in Cardiff, Wales, and Turin, Italy. The primary goal of the CATS study is to determine in a randomized, prospective, controlled study the impact of treating maternal hypothyroidism and hypothyroxinemia on pregnancy and neurodevelopmental outcome of children. The protocol involves collection of blood samples in pregnant women at less than 16 wk gestation, immediate measurement of thyroid function in half the women, and initiation of T₄ therapy in hypothyroid and hypothyroxinemic women. The remaining half have thyroid function tests in the collected pregnancy blood samples measured postpartum, and treatment is then initiated. The treated women and a matched control group without T₄ treatment are being followed for neurodevelopmental parameters in their offspring. Urine samples were randomly chosen from euthyroid and hypothyroid/hypothyroxinemic women for perchlorate and iodine measurements.

The average perchlorate exposure in the women studied by Pearce et al. (15), as determined by urinary perchlorate (median in Turin, 5 μg/liter; Cardiff, 2 μg/liter) (15), was similar to those reported in the study of women in the United States (geometric mean urinary perchlorate, 2.84 μg/liter) (14). In the Pearce study, however, there was no association found between perchlorate exposure and thyroid function. The strength of the study is that the urinary iodine levels were relatively low in both Cardiff (median urinary iodine, 98 μg/liter) and Turin (52 μg/liter). This indicates that even with relatively low iodine intake, perchlorate did not influence maternal thyroid function, at least at the first prenatal visit (<16 wk of pregnancy) and with the levels of perchlorate exposure studied.

More than 1000 women were studied, and statistical analysis showed that the Pearce et al. study (15) was sufficiently powered to detect an effect of the magnitude reported in the study by Blount et al. (14). There were differences in the Cardiff and Turin study sites with respect to the methods used for perchlorate and iodine measurement methods. The free T₄ assay was different between the sites, and the free T₄ by analog method used can be misleading in pregnancy (16). There also was not an accounting in the urinary perchlorate measurement for variations in individual hydration status, such as with a urinary creatinine measurement. Overall, however, these variations in assays and measurements occurred across the study population and should not have been systematically influenced by thyroid function, iodine intake, or perchlorate exposure.

The Pearce study (15) is reassuring evidence that perchlorate exposure at these levels does not alter thyroid function at a susceptible time of pregnancy, even with low iodine intake. Is there still cause for concern? Perchlorate exposure may have effects in pregnancy that are not reflected in maternal thyroid function. Multiple environmental agents can impact thyroid hormone regulation of fetal brain development, and the approaches to detect these effects at an early stage are being explored (17). Perchlorate crosses the placenta and may influence fetal thyroid function, acting directly on the fetal thyroid gland or by influencing iodide transport. Fetal thyroid reserve and capacity for compensation are less than those of the adult. It is possible that fetal exposure to perchlorate impacts fetal thyroid hormone production and brain development, without necessarily altering maternal thyroid hormone levels. Iodine insufficiency could certainly compound these deficits.

The potential for perchlorate effects on the fetus was addressed in a recent study by Blount et al. (18) of 150 pregnant women who were not known to have thyroid dysfunction and were undergoing cesarean section surgery. Maternal urine, maternal serum, cord blood serum, and amniotic fluid were obtained, and perchlorate, nitrate, thiocyanate, and iodine were measured, although not all subjects had all of these measures. Perchlorate was detected in all maternal urine samples and most maternal serum samples (94%). Maternal urine perchlorate levels were always higher than amniotic fluid levels (mean ratio, 22:1), maternal serum perchlorate levels were mostly higher than cord blood levels, and amniotic fluid iodine levels were generally higher than maternal fluid levels. There was no correlation of the level of any NIS inhibitors and outcome measurements including birth weight, length, or head circumference (18). Although the absence of an effect of perchlorate on thyroid function may have been influenced by the high levels of iodine intake in these women, these data indicate that perchlorate is not concentrated in amniotic fluids or fetal circulation and also indicates that measurement of maternal urine perchlorate is a good surrogate marker for fetal perchlorate exposure.

There are multiple sources of perchlorate and multiple inhibitors of iodine uptake that are additive in their effects (7). As these sources are being identified and contamination reduced, the most direct approach to reducing risk of perchlorate exposure in an individual is to ensure adequate iodine intake, especially in the reproductive years for women. This has been advocated in recommendations from the American Thyroid Association (2,19) and The Endocrine Society (10). Maternal thyroid status in pregnancy is likely to be most important for fetal brain development in the first trimester, so women of reproductive age should be especially careful about having an adequate iodine intake, around 250 μg/d (10), for which a supplement of 150 μg/d is recommended (19). The majority of prenatal vitamins do not contain this level of iodine supplement, so it is important to carefully review the content of any supplement being taken (20).

The study by Pearce et al. (15) establishes that in this population of pregnant women in the United Kingdom and Italy with relatively low iodine intake, perchlorate did not impact maternal thyroid function in early pregnancy. Perchlorate may have more subtle effects, and because developmental outcome in children will be performed in the CATS study, it will be useful to consider perchlorate exposure as a variable in these studies. Subsequent investigation may also further define these apparent differences in the influence of perchlorate in studies from the United States and Europe. As these issues are being addressed, women of reproductive age would be best served by ensuring adequate iodine intake, usually requiring a vitamin containing iodine.