Perchlorate Exposure in Pregnancy and Cognitive Outcomes in Children: It's Not Your Mother's Thyroid

Perchlorate (ClO⁻₄) is utilized in manufacturing rocket fuel, fireworks, and airbag inflation systems as well as being formed naturally in the atmosphere (1). It is stable and soluble in water and has been detected in drinking water, vegetables, and even breast milk. Perchlorate is a direct inhibitor of iodine transport and is the best characterized toxicant with respect to mechanism of action, although a range of environmental toxicants have been identified that disrupt thyroid hormone synthesis, metabolism, and action (2). At a sufficient dose and duration of exposure, perchlorate can reduce thyroid hormone production by reducing iodine uptake into the thyroid. Iodine is essential for thyroid hormone production and must be derived from dietary sources (3). Perchlorate antagonism of iodine transport is additive with other inhibitors, such as thiocyanate, found in cigarette smoke. The impact of perchlorate on thyroid function, especially in pregnancy, has led to intense focus on determining acceptable levels of perchlorate and other iodine inhibitors in the water supply (4, 5).

The Controlled Antenatal Thyroid Screening Study (CATS) enrolled 21 846 women at gestational age < 16 weeks from 2002–2006 in Cardiff, United Kingdom, and Turin, Italy (6). Women had blood samples taken, half had thyroid function measured, half of the samples were stored for testing after delivery, and treatment was initiated if the thyroid tests were abnormal. Women in the immediate testing group, whose thyroid tests showed hypothyroidism or isolated hypothyroxinemia (499 or 4.6% of the group), were treated with levothyroxine (6). Neurocognitive outcome of the offspring was evaluated and compared to women in the delayed testing group with hypothyroidism or hypothyroxinemia (551 or 5% of the group), who were not treated until after delivery. Psychological testing at 3 years of age was completed in 390 children (78.2%) with mothers treated during pregnancy and 404 children (73.3%) treated after delivery. The initiation of levothyroxine treatment during pregnancy did not influence neurocognitive outcome, as determined by IQ score in offspring at 3 years of age. The absence of an impact of maternal hypothyroidism on neurocognitive outcome has influenced the continued debate on thyroid function test screening in pregnancy, largely driven on the value of early diagnosis and treatment of maternal hypothyroidism to improve cognitive outcomes (7).

The CATS study subjects were evaluated for the influence of perchlorate on maternal thyroid function (8). Perchlorate was detected in the urine of all women studied, but there was no correlation of perchlorate levels in the urine with maternal thyroid function. This finding was especially significant because the median iodine intake in Cardiff and Turin, 98 and 52 μg/L, respectively, was relatively low. Individuals with lower iodine intake are thought to be more susceptible to the negative effects of perchlorate exposure. In a study of pregnant women in an area of Chile with a high level of perchlorate exposure, there was no impact of perchlorate on maternal or neonate thyroid function, although iodine intake was very high and may have minimized the effects of perchlorate (9). In women from the United States and Argentina in the first trimester of pregnancy with median urinary iodine intake of 144 and 130 μg/L, respectively, perchlorate exposure across a wide range of amounts was not associated with changes in maternal thyroid function (10). In a recent study, 200 pregnant Thai women with urinary perchlorate levels of 1.9 μg/L, thiocyanate levels of 510.5 μg/L, and normal urinary iodine of 153.5 μg/L were studied in the first trimester (11). A positive correlation was found of perchlorate and thiocyanate in the urine with maternal serum TSH concentration and a negative correlation with serum free T₄ concentration. Mild to moderate perchlorate exposure, therefore, has no effect or modest effects on thyroid function, and sufficient iodine intake generally mitigates the influence on thyroid function tests, although this has not been consistent in all studies. Despite the absence of consistent alterations in maternal thyroid function in most perchlorate exposure studies, the potential for other actions of perchlorate on the fetal thyroid or direct effects on the brain have been proposed (2, 5).

The study by Pearce and colleagues (12) in this issue of the JCEM, utilizing data from the CATS study, addresses the important question of linking maternal perchlorate exposure in pregnancy to cognitive outcome of their offspring. The CATS study identified 1050 pregnant women with hypothyroidism or hypothyroxinemia; half were in the immediate treatment group, and half were in the group tested and treated after pregnancy. In the combined screening and control groups, 794 (75.6%) children completed the psychological testing, and 487 (46.4%) mother-child pairs had completed psychological testing and urinary iodine and perchlorate measurements. The 487 women-child pairs in this study, therefore, represent approximately two-thirds of those reported in the study of T₄ treatment effects on cognitive outcome (6) and included women treated and not treated with T₄ during pregnancy. Women with urinary perchlorate concentrations in the upper 10% had a significantly increased risk of having offspring with a reduced verbal IQ at 3 years of age (12). There was no association of perchlorate levels with performance IQ. The association with verbal IQ was not influenced by maternal thyroid status and, surprisingly, not influenced by maternal administration of T₄, which about half of the women received as part of their group assignment. Interestingly, a recent study from the United Kingdom showed that women with low urinary iodine in the first trimester had an increased risk of offspring with a lower verbal IQ, but not performance IQ, compared to the control group (13). The strength of the current study is linking maternal exposure of a thyroid toxicant in pregnancy directly to a cognitive outcome. The absence of a direct effect of perchlorate on maternal thyroid function (8), however, focuses interest back to developmental effects of perchlorate that are not necessarily reflected in maternal thyroid hormone levels.

The sodium-iodide symporter is expressed in the placenta and can transport perchlorate. Direct measurement of perchlorate in the amniotic fluid and cord blood shows detectable levels, although perchlorate is not concentrated or found at levels higher than those in the urine (14). Several animal studies have shown direct effects of perchlorate on brain development (2). A study in rats showed reduced hippocampal synaptic transmission in adult offspring of perchlorate-exposed pregnant dams, which occurred in a dose-dependent manner (15). At some doses, TSH was elevated and T₄ was reduced, but serum T₃ remained normal. Perchlorate may also influence the impact of other environmental toxicants. A recent study in zebrafish showed that the addition of perchlorate to BDE-47, a polybrominated diphenyl ether, resulted in greater disruption of thyroid hormone-dependent gene expression in the brain (16). Rodent models also indicate that the metabolic compensation that should occur in the brain when the serum T₄ is low (increased T₄ to T₃ conversion with a falling T₄) may not be sufficient to maintain baseline gene expression in all areas of the brain (17).

The direct effect of perchlorate on fetal thyroid gland function is another possible mechanism for perchlorate to influence fetal brain development. The fetal thyroid gland has less stored reserve than the adult and is likely more susceptible to iodine deprivation, either directly or due to perchlorate inhibition. Measurement of urinary goitrogens in infants, including perchlorate, showed an association of the goitrogen levels with an increase in urinary TSH among those infants with lower iodine intake (18). If maternal thyroid function is adequate, as it was at least for women receiving T₄ replacement in the Pearce study (12), we generally believe that this should be sufficient to promote normal brain development in the fetus. An example is congenital hypothyroidism, when the fetus is not able to synthesize its own thyroid hormone and relies completely on maternal T₄ crossing the placenta (19). Infants with congenital hypothyroidism, when promptly diagnosed by screening and appropriately replaced with T₄, generally experience normal brain development. There are, however, reports of specific deficits in sensory development, sensory processing, and in some studies, measured IQ, despite early and adequate therapy after birth (19).

The recommendation to ensure adequate iodine intake during pregnancy, including iodine supplementation, has become more widespread and was recently endorsed by the American Academy of Pediatrics (20). This remains sound advice and is endorsed in pregnancy guidelines by both The Endocrine Society (21) and the American Thyroid Association (22). With respect to the impact of perchlorate and other sodium-iodide symporter inhibitors, such as thiocyanate contained in cigarette smoke, on maternal thyroid function, adequate iodine intake remains the best approach to mitigate the impact of these thyroid toxicants.

The finding that maternal perchlorate exposure insufficient to influence maternal thyroid function was associated with adverse cognitive outcome in the offspring (12), however, points to a significantly broader impact of perchlorate on brain development. The cornerstone of the recommendations from the National Academy of Science Panel on Perchlorate Contamination was that the first detectable action of perchlorate is inhibiting iodine uptake, and the first adverse effect is when iodine inhibition is sufficient to reduce thyroid hormone production (4). The findings from the current study challenge this view and should stimulate further investigation of the direct effects of perchlorate on brain development. Although urinary perchlorate levels appear to be a good surrogate for fetal exposure to perchlorate, measurement of perchlorate in amniotic fluid may be a more direct way to assess fetal exposure (14). It is especially important to consider the potential for interaction among different toxicants.

Studies that determine the impact of toxicant exposure, such as perchlorate, on cognitive outcomes are the most important to perform, but they are also the most difficult due to the long-term nature of follow-up required as well as the subtle nature of the impact of toxicants like perchlorate. The potential mechanisms include effects on maternal and fetal thyroid function, direct actions on brain development, and finally, perchlorate levels as a surrogate for exposure to other environmental toxicants. Adequate iodine nutrition remains the best approach to mitigate the impact of perchlorate. Mechanistically, studies of adaptation to low iodine in the fetus, regulation of thyroid hormone action and developmental windows, as well as direct actions of environmental toxicants, acting as both agonists and antagonists, are important. This study by Pearce and colleagues (12), however, suggests that adequate iodine nutrition alone may not be enough to protect the developing fetus, and more aggressive efforts to lower environmental perchlorate exposure may be indicated, especially if larger studies confirm these findings.