Thyroid Disruption and Exposure
How perchlorate from rocket fuel, fertilizers, and contaminated water blocks thyroid hormone production — and who is most at risk
Perchlorate is a chemical found in rocket fuel, explosives, and certain fertilizers that has contaminated drinking water and food supplies across the United States and worldwide. It blocks the thyroid gland's ability to absorb iodine — the essential raw material for thyroid hormones — by competing with iodide at the same transport site. A large national survey detected perchlorate in every urine sample tested, meaning virtually everyone is exposed [1]. The concern is greatest for pregnant women, infants, and anyone with borderline low iodine intake, because even modest thyroid disruption during fetal development can affect brain development.
How Perchlorate Disrupts the Thyroid
The thyroid gland concentrates iodine from the bloodstream using a protein called the sodium-iodide symporter (NIS). Perchlorate binds to NIS with higher affinity than iodide itself, effectively blocking iodine uptake. With less iodine entering the thyroid, production of T3 and T4 — the hormones that regulate metabolism, energy, and development — falls. The pituitary gland compensates by releasing more thyroid-stimulating hormone (TSH), which is why elevated TSH on a blood test can signal perchlorate burden alongside other causes.
The key moderating factor is iodine status. When dietary iodine is plentiful, the competition from perchlorate is partially overcome by sheer quantity of iodide. The 2006 NHANES study of 2,299 Americans found that in women with low urinary iodine, perchlorate significantly predicted lower T4 and higher TSH — but effects were much weaker in women with adequate iodine intake [1]. This is why maintaining sufficient iodine is the most practical protective step.
Perchlorate does not act alone. Nitrate (from well water, vegetables, and processed meats) and thiocyanate (from cruciferous vegetables and cigarette smoke) block the same NIS transporter. Their combined burden is more predictive of thyroid disruption than any single chemical measured in isolation [5].
Where Exposure Comes From
Water: Industrial contamination is widespread. Rocket testing and production facilities have contaminated groundwater in Nevada, California, Texas, and other states. The Colorado River — drinking water for Los Angeles and San Diego — was contaminated from a Henderson, Nevada facility. Roughly 70% of US military installation groundwater samples tested between 1997 and 2009 contained measurable perchlorate. The EPA has proposed a drinking water limit of 20 µg/L (the rule has been in progress since 2011 and is expected to be finalized in 2027).
Food: An FDA survey found detectable perchlorate in at least one sample of 74% of common foods tested. The highest concentrations appear in dairy products, leafy greens, and some root vegetables, because plants absorb perchlorate through their roots from contaminated soil and irrigation water [6]. Chilean nitrate fertilizers used in agriculture are a natural source of perchlorate that has spread into agricultural soils globally.
Infant formula: All types of commercially produced formula contain measurable perchlorate — detected in 100% of samples in multiple studies. Cow's milk-based formula tends to carry higher concentrations than soy-based varieties. When formula is prepared with tap water that is also contaminated, exposures can exceed the EPA reference dose [4].
Breast milk: Perchlorate was detected in 100% of breast milk samples in a Boston-area study, with concentrations ranging widely (1.3 to 411 µg/L). About half of the nursing women in that study appeared to be providing breast milk with potentially insufficient iodine — a compounding problem, since the infant faces both lower iodine supply and active competition from perchlorate [3].
Who Is Most Vulnerable
Pregnant women and fetuses are the highest-priority concern. The fetal brain depends entirely on maternal thyroid hormone through the first half of pregnancy, before the fetal thyroid is functional. Even modest reductions in maternal T4 during this window have been associated with lower IQ, delayed motor development, and attention deficits in children. The Steinmaus 2016 study of nearly 1,900 pregnant women in Southern California found perchlorate exposure associated with meaningful decreases in both total T4 (β = −0.70) and free T4 (β = −0.053), and increased TSH [2].
Formula-fed infants face exposure from both the formula itself and any perchlorate in the water used to prepare it. Infants have high metabolic demands for thyroid hormone relative to their body weight and less iodine reserve than adults.
People with low iodine intake — including those avoiding iodized salt, dairy, and seafood — have less buffer against NIS competition. Iodine deficiency and perchlorate exposure together create a worse outcome than either alone [5].
Practical Steps to Reduce Exposure
Water filtration: Reverse osmosis (RO) systems effectively remove perchlorate from drinking water and are the most practical point-of-use option. Standard activated carbon filters — pitcher filters, most faucet filters — do not remove perchlorate. This distinction matters especially when preparing infant formula.
Iodine sufficiency: The most evidence-backed protective factor is maintaining adequate iodine intake. Sources include iodized salt, seaweed (kelp, nori), dairy, eggs, and seafood. The RDA for iodine is 150 µg/day for adults, rising to 220 µg during pregnancy and 290 µg while breastfeeding. Many prenatal vitamins do not include iodine, or include too little — check the label.
Produce: Washing vegetables reduces surface-level contamination but does not remove perchlorate absorbed into plant tissue. Buying organic reduces some pesticide and fertilizer-source contamination, but naturally occurring perchlorate in soil means organic is not entirely protective.
See our water filtration page for a broader look at what's in tap water. For more on the thyroid, see our thyroid health pages.
Evidence Review
Population Exposure: NHANES 2001–2002
The landmark Blount et al. (2006) study analyzed urine from 2,299 participants in the National Health and Nutrition Examination Survey (NHANES), the most comprehensive US population sample [1]. Perchlorate was detected in all 2,820 urine specimens — establishing universal background exposure across the American population.
The thyroid hormone findings showed a clear pattern of sex and iodine-status modification. In women with low urinary iodine (below 100 µg/L), higher urinary perchlorate predicted both decreased total T4 (p < 0.0001) and increased TSH (p = 0.001). In women with adequate iodine, perchlorate still predicted increased TSH (p = 0.025) but effects were attenuated. No significant associations were found in men — a difference attributed to women's typically lower body mass (higher dose per kg) and iodine intakes generally closer to the deficiency threshold. The authors noted that effects were "observed at perchlorate exposure levels that were unanticipated based on previous occupational studies."
Pregnancy Cohort: Southern California
Steinmaus et al. (2016) followed 1,880 pregnant women in San Diego County who were exposed to drinking water contaminated from an industrial source, with median urinary perchlorate approximately twice the general US population median [2]. Adjusting for confounders, higher perchlorate was associated with:
- Decreased total T4: β = −0.70 µg/dL (95% CI: −1.06, −0.34)
- Decreased free T4: β = −0.053 ng/dL (95% CI: −0.092, −0.013)
- Increased TSH: β = 0.071 mIU/L (95% CI: 0.008, 0.133)
These effect sizes are clinically meaningful during pregnancy. The study design was prospective and the contamination source industrial, making this a real-world exposure scenario rather than a laboratory model.
Breast Milk and Infant Formula
Pearce et al. (2007) detected perchlorate in 100% of 49 breast milk samples and 100% of 56 urine samples from Boston-area lactating women [3]. Perchlorate concentrations in breast milk ranged from 1.3 to 411 µg/L — a 300-fold range indicating wide individual variation in exposure and excretion. Crucially, about 47% of sampled women appeared to be providing breast milk with potentially insufficient iodine for infant needs. Because perchlorate concentration was not significantly correlated with breast milk iodine concentration, the two problems are largely independent: an infant can face both low iodine delivery and active NIS inhibition simultaneously.
Schier et al. (2010) specifically examined perchlorate in commercial infant formulas [4]. Cow's milk-based formulas carried the highest concentrations (mean 1.72 µg/L; range 0.68–5.05 µg/L), while soy-based formulas averaged 0.21 µg/L. When cow's milk formula is prepared with tap water containing even moderate perchlorate contamination, estimated daily infant intake can exceed the EPA reference dose of 0.0007 mg/kg/day.
Infants: Compounding Exposures
Cao et al. (2010) followed 92 full-term infants through their first year of life, collecting 206 urine samples [5]. TSH was significantly elevated in infants with low iodide combined with high perchlorate. The effect largely disappeared when iodide status was adequate — consistent with the mechanistic model. Importantly, this study also measured thiocyanate and nitrate, finding that combined NIS-inhibitor burden (all three goitrogenic anions together) was more predictive of TSH elevation than perchlorate alone. This cumulative burden model better reflects real-world infant exposure, where multiple NIS inhibitors coexist.
Dietary and Environmental Sources
Steinmaus (2016) reviewed population exposure data and FDA food surveys [6]. The FDA 2005–2006 survey of 280 foods found detectable perchlorate in at least one sample of 74% of foods tested. Highest concentrations appeared in dairy (sour cream, milk, ice cream), cured meats, and certain vegetables including collards, tomatoes, and potatoes. Natural perchlorate in Chilean nitrate fertilizers routes the chemical into agricultural soil globally, meaning contamination is not confined to industrial zones.
Strength of Evidence and Limitations
The mechanistic evidence (NIS inhibition) is well-established across multiple independent research groups. The population association evidence from NHANES and the Southern California pregnancy cohort is strong for thyroid hormone effects in exposed women, with effect sizes consistent across studies. Limitations include the observational design in human studies (cannot randomize perchlorate exposure), difficulty separating perchlorate from co-occurring NIS inhibitors, and the challenge of attributing health outcomes beyond thyroid hormones (such as cognitive outcomes in children) to maternal perchlorate specifically. The overall weight of evidence was sufficient for the EPA to determine in 2011 that perchlorate poses a meaningful public health risk warranting a drinking water regulation — a position reaffirmed by the courts when that determination was reversed in 2020 and then reinstated in 2023.
References
- Urinary Perchlorate and Thyroid Hormone Levels in Adolescent and Adult Men and Women Living in the United StatesBlount BC, Pirkle JL, Osterloh JD, Valentin-Blasini L, Caldwell KL. Environmental Health Perspectives, 2006. PubMed 17185277 →
- Thyroid Hormones and Moderate Exposure to Perchlorate during Pregnancy in Women in Southern CaliforniaSteinmaus C, Pearl M, Kharrazi M. Environmental Health Perspectives, 2016. PubMed 26485730 →
- Breast milk iodine and perchlorate concentrations in lactating Boston-area womenPearce EN, Leung AM, Blount BC. Journal of Clinical Endocrinology and Metabolism, 2007. PubMed 17311853 →
- Perchlorate exposure from infant formula and comparisons with the perchlorate reference doseSchier JG, Wolkin AF, Valentin-Blasini L. Journal of Exposure Science and Environmental Epidemiology, 2010. PubMed 19293845 →
- Goitrogenic Anions, Thyroid-Stimulating Hormone, and Thyroid Hormone in InfantsCao Y, Blount BC, Valentin-Blasini L. Environmental Health Perspectives, 2010. PubMed 20439182 →
- Perchlorate in Water Supplies: Sources, Exposures, and Health EffectsSteinmaus CM. Current Environmental Health Reports, 2016. PubMed 27026358 →
Transparency
View edit historyEvery change to this page is tracked in version control. If you have conflicting research or think something is wrong, we want to hear about it.