Perchlorate salts are used as oxidizers in solid propellants for rockets and missiles, fireworks, road flares, matches, and air bag inflation systems. Perchlorate is also in Chilean nitrate fertilizers exported for use around the world. Low levels of perchlorate may also be found in the environment due to natural processes (11). Perchlorate has been detected in the drinking water of communities around the United States and worldwide. Since perchlorate is essentially not biodegradable, industrial contamination of water supplies is difficult and expensive to remove. Perchlorate has also been detected in foods such as lettuce and other produce, wheat, cows’ milk, wine and beer, and multivitamins (including prenatal multivitamins). Most US perchlorate exposure likely occurs through the diet rather than the water supply (12). Perchlorate exposure appears to be ubiquitous. In the US population, perchlorate was detectable in all 2,820 spot urine specimens from the 2001–2002 National Health and Nutrition Examination Survey (NHANES) with a median urine concentration of 3.6 μg/L (13). Among the 36% of women with urinary iodine values <100 μg/L in the NHANES data set, there was a significant positive correlation between urinary perchlorate concentrations and serum TSH and an inverse correlation between urinary perchlorate concentrations and serum T₄ values (14). Among women with urinary iodine values >100 μg/L, there was a positive association between urinary perchlorate concentrations and serum TSH, but there was no association between urinary perchlorate concentrations and serum T₄ values. There were no significant associations between urinary perchlorate concentrations and thyroid function values in men, irrespective of their urinary iodine concentrations.

However, the administration of 3 mg perchlorate to normal volunteers throughout the day for 14 days did not affect serum TSH, thyroid hormone concentrations, or the 24-hour thyroid ¹²³I uptake (15). In a second study, 10 mg perchlorate daily for 14 days also had no effect on thyroid function tests in spite of a 50% decrease in the thyroid ¹²³I uptake (16). Volunteers who received doses ranging from 0.5 to 35 mg perchlorate daily for 14 days had no change in serum TSH, T₄, or T₃ concentrations; there was a slight decrease in the thyroid ¹²³I uptake at the 1.4 mg and higher doses, but not at the 0.5 mg dose (17). Normal volunteers who received daily doses of 0.5 or 3.0 mg perchlorate for 6 months had no changes in serum TSH, thyroglobulin, T₄, FT₄ index, total T₃, and the thyroid ¹²³I uptake, although this study was underpowered (18).

An occupational study examined 29 men working in a factory producing ammonium perchlorate who had been exposed to high-level airborne perchlorate in 12-hour shifts 3 to 4 days/week for an average of 3 years (19). Thyroid function was normal and did not differ at the end of three 12-hour shifts compared to their 3 days away from the plant, apart from a decrease in the 14-hour thyroid ¹²³I uptakes. Thyroid ultrasounds did not reveal any significant abnormalities, and thyroid volumes were similar to 12 unexposed community controls.

The developing fetus is likely to be most vulnerable to the adverse effects of perchlorate exposure since thyroidal iodine turnover is highest in fetal life and the fetus requires adequate thyroid hormone for normal neurodevelopment. A Chilean study found that thyroid hormone and TSH levels in pregnant
women and their neonates did not vary among three towns with drinking water perchlorate concentrations of 0.5 μg/L, 6 μg/L, and 114 μg/L (20). However, the women in all three towns had high urinary iodine concentrations (median 269 μg/L), which may have made them less vulnerable to the adverse effects of perchlorate. In Israel, neonatal T₄ levels did not differ in the infants of women who had consumed drinking water with high compared to low perchlorate concentrations during pregnancy (21). An ecologic study in California found increased neonatal TSH levels in communities with drinking water perchlorate concentrations >5 μg/L compared to communities with lower drinking water concentrations (22). However, there were no correlations between urinary perchlorate concentrations and first-trimester serum free T₄ or TSH values in a group of 1,641 pregnant women from Cardiff, Wales and Turin, Italy, including the majority who had urinary iodine concentrations <100 μg/L (23). Similarly, no associations were detected between urine perchlorate concentrations and thyroid function among 134 women from Los Angeles and 107 pregnant women from Córdoba, Argentina (24).

Iodine, required for infant nutrition, is secreted into breast milk via NIS (25), and this process may be inhibited by perchlorate. A recent study measured perchlorate in the breast milk of women from around the US and found detectable levels in all 36 samples (range 0.6 to 92.2 μg/L) (26). Breast milk iodide and perchlorate concentrations were inversely correlated in the six samples with perchlorate concentrations ≥10 μg/L, although there were no correlations between breast milk iodide and perchlorate in the full data set. In a study of 56 Boston-area lactating women, perchlorate was detectable in all breast milk samples (median, 9.1 μg/L) (27). However, there were no significant correlations between breast milk iodine and perchlorate concentrations, including in the subset of 23 women whose breast milk perchlorate concentrations were ≥10 μg/L. It has been shown that perchlorate itself is transported into breast milk via NIS (28). In a study of 92 full-term infants, higher urinary perchlorate, nitrate or thiocyanate was associated with higher urinary TSH values (29). However, the correlation between urine and serum TSH was relatively poor (Spearman correlation coefficient 0.49) in the subset of 50 infants for whom serum testing could be performed and there were no significant associations between perchlorate and serum thyroid function tests in those 50 infants. A preliminary study in 64 mother–infant pairs found no associations between maternal breast milk or infant urine iodine concentrations and infant thyroid function at 2 months of age (30). Overall, the effects of low-level environmental perchlorate exposure on thyroid function remain unresolved.

Thiocyanate is about 15 times less potent than perchlorate as an NIS competitor (31). Cigarette smoke contains cyanide that is metabolized to thiocyanate. Thiocyanate is also a metabolite of cyagenic glucosides present in plant foods such as cassava, cabbage, turnips, broccoli, brussels sprouts, and cauliflower. It has been reported that women who smoke during pregnancy are more likely to give birth to neonates with decreased serum T₄ levels, increased TSH levels, and thyroid enlargement (32,33). Among Boston-area women in the first trimester of pregnancy, FT₄I levels were lower in women who reported current smoking than they were in non-smokers (34). Other studies concluded that cigarette smoking decreases breast milk iodine concentrations (35,36). These effects are postulated to be related to thiocyanate exposure. It has also been shown that diets high in thiocyanate contribute to the development of goiter in iodine deficient regions (37). However, overall the effects of cigarette smoke on thyroid function appear to be complex and somewhat contradictory. Large population studies have demonstrated lower serum TSH levels (38) and higher free T₄ and free T₃ levels (38,39) in smokers than nonsmokers; the mechanism is unknown.

Nitrate is about 240 times less potent than perchlorate as a competitive inhibitor of NIS (31). Nitrates occur naturally in soil and groundwater due to the decomposition of organic materials, and are present in plants. Sodium nitrite is also used as a preservative in cured meats and other foods. In the United States, the average daily dietary intake of nitrate in adults is about 75 to 100 mg, with about 80% to 90% from vegetables such as beets, celery, lettuce, and spinach (40). Vegetarians may have nitrate intakes of up to 250 mg/day. In addition, inorganic nitrates are used as fertilizers. High nitrate ingestion may occur due to contamination of the water supply by human sewage or livestock manure, especially from feedlots, or by runoff from farmland. The World Health Organization has defined the safe upper limit for nitrate in drinking water as 50 mg/L (41) and the U.S. Environmental Protection Agency (EPA) has set a maximum contaminant level at 45 mg/L (42); these guidelines were adopted to protect against methemoglobinemia in bottle-fed infants rather than because of concerns about thyroid health.

Among 3772 iodine-sufficient adults in the Study of Health in Pomerania, there was no cross-sectional association between urinary nitrate concentrations and thyroid volume by ultrasound (43). Similarly, in 10 healthy volunteers, no changes in thyroid function tests or radioactive iodine uptake were found following 4 weeks of the daily ingestion of 15 mg/kg sodium nitrate, far above the nitrate concentration typically ingested in the US diet (44). Recent studies have described the effects of very high nitrate exposures due to drinking water contamination in areas of Eastern Europe. Among iodine-sufficient schoolchildren in a Bulgarian village where drinking water nitrate levels were 75 mg/L, the odds ratio for goiter was 3.01 compared to children in a nearby village where drinking water nitrate levels were only 8 mg/L (45). Another Bulgarian study found an odds ratio of 5.3 for thyroid dysfunction among pregnant women living in a village with high drinking water nitrate content (93 mg/L) compared to women in a village without high nitrate exposure (46). Finally, sonographic thyroid volumes in iodine-sufficient Slovakian schoolchildren from a region with high drinking water nitrate levels (51 to 274 mg/L) and the prevalence of subclinical hypothyroidism were significantly higher compared to age-matched children without high nitrate exposure (47).

Two recent studies have suggested an association between high nitrate ingestion and thyroid cancer. An ecologic study in 21,977 Iowa women found no effect of drinking water nitrate concentrations on thyroid function, but did find an increased relative risk (2.6; 95% CI 1.1 to 6.2) for thyroid cancer with higher nitrate exposures (48). Among 490,194 participants
in the NIH-AARP Diet and Health Study, increased thyroid cancer risk was observed in men, but not women, with higher self-reported dietary nitrate intakes (49).

Isopropylidenediphenol, or bisphenol-A (BPA) has been shown to bind to the thyroid hormone receptor and antagonize its activation (73). BPA is a plastic monomer used to produce polycarbonate plastic. Over 6 billion pounds of BPA are produced and over 100 tons released into the atmosphere worldwide annually (74). It is used in food containers, baby bottles, and reusable water bottles, and is found in epoxy resin linings of some metal food cans. It may leach from these containers into stored food and drink. BPA is also used in dental sealants. Studies from developed countries around the world have found BPA in human serum, urine, amniotic fluid, and breast milk (75). In a reference population of 394 U.S. adults, BPA was detected in 95% of urine samples with a median concentration of 1.28 μg/L (76). The U.S. EPA has defined the reference dose for BPA as 50 μg/kg/day. The European Union has set a no-observed-adverse-effect level (NOAEL) of 5 mg/kg/day.