Perchlorate – properties, toxicity and human health effects: an updated review

Anthropogenic

Perchlorate was described for the first time by Count Friedrich von Stadion in the work “Von den Verbindungen des Chlorine mit dem Sauerstoff” in 1816 [13]. Significant increase in worldwide production of perchlorate was observed during World War II. In 1940 global production of ammonium perchlorate was about 2,000 tons per year, then in 1945 it was approximately 20,000 tons [14]. In this period perchlorate were mainly used for the production of explosives for war industry. After 1945, the production of perchlorate continued to grow especially for military applications, as their interesting properties were discovered as a valuable component of rocket fuel [15]. In conjunction with the current use of perchlorate in military installations, data on production sites and quantities of perchlorate production are only partially shown publicized. Thus, information on the annual production of perchlorate are inherently incomplete and consequently far from being actual. It is estimated that the annual global production of perchlorate is about several hundred thousand tons and the largest production plants are placed in the USA, France, Germany, Italy, China and Brazil [16], [17].

Anthropogenic sources include the import and use of Chilean nitrate as a fertilizer exploited from deposits in the Atacama desert in Chile. The fertilizer produced on the basis of Chilean nitrate contains approximately 98% NaNO₃ and, according to various studies, from 0.05 to 0.4% ClO₄⁻ [18]. In the 1830–1980 period, the estimated volume of imports of Chilean saltpeter ore to the USA and Europe amounted to approximately 23 million tons in terms of nitrogen, with the highest value being reported in 1930s [19]. In US commercial production of sodium nitrate began in 1915 and in the following years domestic production increased its share in total US nitrate consumption. In 1994 over 11 million tons of sodium nitrate were used in US agriculture and only approximately 30,000 tons were from imported Chilean nitrate [2]. According to United Nations Trade Statistics Branch the volume of sodium nitrate import from Chile was about 19,000 tons in 2018 [20]. A material evidence of increasing man-made perchlorate contamination has been disclosed in shallow ice core in Thienshan, East Asia, where concentration of ClO₄⁻ was still increasing during 1980s’ to 2004 [21].

Natural

The largest known perchlorate reservoir in the world are the aforementioned Chilean saltpeter deposits. The presence of natural perchlorate has also been demonstrated in snow samples from Arctic, waters of the Great Lakes (USA and Canada), soil samples from Death Valley (USA) and Dry Valley (Antarctica) [1], [22], [23], [24], [25]. In 2015, the results of determination of perchlorate content in surface and underground waters as well as soil samples from semi-arid areas and deserts (Atacama Desert in Chile, Mojave Desert in the USA, deserts in western regions of Africa), showed the presence of ClO₄⁻ in range from 10⁻¹ to 10⁶ μg/kg [26]. The mechanism of formation of perchlorate in nature is not fully understood. Several paths of a possible perchlorate natural formation have been proposed [2], [26], [27], [28], [29], [30], [31], mainly in photochemical processes, with the participation of chloride ions and ozone in the troposphere and stratosphere [32]. The theory indicating the possibility of perchlorate formation during atmospheric discharges [2], [33] has not been explicitly confirmed [16]. It was also shown that the isotopic composition of perchlorate of natural origin (e.g., originating from Chilean nitrate) is different from those made artificially, which facilitates identification and determination of the source in the environment [34], [35].

Physical and chemical properties

Owing to the fact that perchlorate was discovered in 19th century, its chemistry seems to be well known [14], [37]. In this review article we gather the most important data on perchlorate properties in order to better understand its impact on human health. ClO₄⁻ is a simple inorganic anion with strong oxidizing properties. Perchloric acid is one of the strongest acids known and it is completely dissociated in water [38]. It has the structure of a tetrahedron composed of four oxygen atoms and one centrally located chlorine atom [14]. The average length of Cl-O bonds was determined experimentally by X-ray crystallography at 1.42 Å [39], [40]. It owes its strongly oxidizing character to the chlorine atom, located on + VII state of oxidation, however, the redox reaction rate involving perchlorate is very small and usually observed only in concentrated perchloric acid solution. This is probably due to the presence of oxygen atoms in a tetrahedral molecule that protects the reactive chlorine atom from the influence of reducers from environment [4]. For the same reason, perchlorate salts are poor complexing agents with metal ions, therefore they are widely used in analytical chemistry [14]. The hydration energy of perchlorate is among the lowest in popular inorganic anions which, in turn, has significant consequences in ClO₄⁻ determination methods, especially ion chromatography [37]. In acid solutions with concentrations of 0.1–4 M no reduction of perchlorate ions is observed by common reducing agents like thiosulphates, sulphides and iron (II) ions. As a result, perchlorate can accumulate in the environment for many years, undergoing very slow degradation [37].

Solid perchlorate are white or colorless crystals, hydrated (sodium perchlorate) or anhydrous (potassium perchlorate). Perchloric acid is a colorless, oily liquid. The solubility of perchlorate in water depends on the type of cation [1]. Most of them are well soluble in water and non-aqueous solvents, owing to their low affinity for metal cations [17]. Basic physical properties of the most commonly used perchlorates are shown in Table 2.

Formation of perchlorate in drinking water systems

Perchlorate prevalence in drinking water may be caused by chlorine or gaseous ClO₂ which are commonly used to disinfect water. The thermodynamically favorable reaction of perchlorate formation is given below [37]:

However, the reaction still lacks scientific evidence. The mechanism of reaction is probably more complex and requires intermediate steps [37]. An evidence of ClO₄⁻ formation during electrochemical oxidation of water was described [41]. Perchlorate would be generated in household bleaches, which contains up to 15% of sodium hypochlorite, NaOCl by disproportion in two-step reaction [37]:

Although the most likely potential pathways of perchlorate formation in drinking water have been proposed, little is known about contribution of ClO₄⁻ from the aforementioned sources [42].

In particular the reactions requires alkaline pH and they do not occur in ambient temperature very fast [37]. On the other hand, the electrochemical water treatment may lead to perchlorate contamination as by-product. There is a report [43] of concentrations up to 9 mg/L ClO₄⁻ obtained in disinfected water, much higher than drinking water equivalent level of reference dose (24.5 μg/L) [44].

Methods of production

Perchloric acid was first obtained in 1816 by von Stadion as a result of the low pressure distillation of mixture of concentrated sulfuric acid and potassium chlorate. In 1818 the same author described the method of obtaining HClO₄ as a result of saturated aqueous chlorine dioxide solution [13], [14]. In both cases a solution of approximately 70% was obtained due to azeothropic mixture forming [14]. Current industrial production methods are based on the chlorates electrolysis to perchlorate. The basic substrates for the production are gaseous chlorine and caustic soda (NaOH) (1). Then hypochlorous acid (HClO) is formed and sodium hydroxide solution used in excess reacts with the obtained acid to give its sodium salt (2). Sodium hypochlorite then undergoes an electrolysis process to form NaClO₃ (3) which, due to further electrolytic oxidation, gives NH₄ClO₄ (4). At last, followed by reaction with sulfuric acid or hydrochloric acid, HClO₄ is formed (5) [45], [46]. The above mentioned reactions (1)–(5) are provided below.

Methods of determination

There are numerous techniques for determining perchlorate, although most of them are currently insufficient due to their relatively high limit of detection (LOD) and limit of quantification (LOQ). They include gravimetric, colorimetric, volumetric, spectrophotometric, potentiometric, capillary electrophoresis, yet ion chromatography methods seems to be most popular now, especially coupled with mass spectrometry detection. Due to good perchlorate solubility gravimetric methods appear to be problematic, yet there is some scientific data on this technique. Perchlorate can be determined in the complex with nitrone or tetraphenylphosphonium perchlorate [47], [48] or after reduction to chloride ions in form of silver chloride [49]. Detection limit of these methods is about 10 mg/L, so the sensitivity is too low for current applications and needs. Determination of perchlorate as a complex form with some colorants such as methylene blue, neutral red etc. is also used in colorimetric [50] and spectrophotometric [51], [52], [53], [54], [55] methods. However, there are many interfering factors which encumber the application of such methods in biological samples (e.g., body fluids). The range of detection limit of these methods is from 20 to 2 mg/L. There are some reports of volumetric determination of perchlorate, with titration of exceed perchlorate reducing agent like Fe²⁺ or Ti³⁺ [56], [57], [58]. There is a report [59] of a sensitive stripping voltammetric method based on voltammetric ion - selective electrode, with detection limit of 0.02 μg/L. A comprehensive review of perchlorate determination methods has been published 10 years ago [60]. Most of the current applications of perchlorate determination in various environmental and clinical samples are based on ion chromatography. The first modern perchlorate determination using ion chromatography (IC) were described in 1997 [61], [62]. IC has many advantages including low detection limits and determination at μg/L and requires only small sample volumes. There are many methods of sample preparation adapted for IC and the possibility of using different detectors. Application of tandem mass spectrometer as a detector results in improving selectivity and sensitivity of the ion chromatography by 100–1,000 times, with the reported detection limit of 0.5 ng/L [63]. IC and IC-MS/MS methods are extensively described elsewhere [64]. For the purposes of this review only the US Environmental Protection Agency (US EPA) methods are presented (see Table 3).

Methods of remediation of perchlorate from the environment

Since the discovery of high concentrations of ClO₄⁻ in the waters of the Colorado River, USA in the late 1990s [1], many studies were conducted to find a close relationship between the presence of these compounds in the environment with human activity (industrial production, import of Chilean nitrate). In recent years, a large amount of work focuses on the natural occurrence of perchlorate in the environment. In connection with the appearance of new reports [69], [70], [71], [72], [73] on the occurrence of perchlorate in water samples, soils, vegetables, dairy products and other food products, there is a need to degrade the ClO₄⁻ anion to other non-toxic substances or to physically remove it from the environment. Due to high mobility of ClO₄⁻ in aqueous solutions, as well as its low reactivity, caused by high energy of activation and weak complexing properties, it is necessary to develop methods for decontamination of perchlorate on a large scale, because the use of conventional techniques such as coagulation, sedimentation and precipitation is expensive and inefficient [37]. In order to reduce or eliminate the threat caused by the presence of perchlorate in the environment, modern remediation methods are developed and improved.

Physicochemical methods include filtration, membrane processes such as electrodialysis, reverse osmosis, ultrafiltration and nanofiltration, adsorption on granular activated carbon, chemical and electrochemical reduction and, above all, ion exchange. Note, that only the last technique is used on a larger scale [74], in particular, for the removal of perchlorate from drinking water. The resins used to exchange perchlorate are essentially composed of a positively charged polymer (e.g., quaternary ammonium compounds) and related anions (e.g., chloride ions), which during the washing of the resin go into solution and ClO₄⁻ ions are retained on the resin surface. It has been shown that for the efficiency of the entire ion exchange process, the balance between kinetics and process selectivity is of great importance. Two-functional resins were developed in which two different quaternary ammonium compounds were used, differing in the chain length of the substituents. The longer chain caused selectivity, while the shorter one was responsible for improving the reaction kinetics. A method of regeneration of such resins was also developed, using FeCl₄⁻ ions obtaining almost 100% recovery, additionally consuming very small amounts of water [37], [75]. Other techniques are complicated to use, require expensive equipment and cause a large amount of waste that is difficult to dispose of [75].

For remediation of perchlorate by biological methods, selected species and microbial strains are used, which in the presence of a suitable electron acceptor (e.g., ClO₄⁻, NO₃⁻, O₂) and carbon sources, can enzymatically convert perchlorate to other, non-toxic compounds, thus reducing the level of harmful substances to a level acceptable for drinking water. In general, perchlorate reducing bacteria can be divided into four main groups: (1) perchlorate reducting bacteria (PRB), which can reduce both chlorates and perchlorate. They belong to the genus Dechloromonas and Azospira. It is a group currently of the greatest importance in the processes of biological remediation of perchlorate. (2) bacteria that reduce chlorate (chlorate reducting bacteria, CRB), having enzymes that only reduce chlorates. Representatives are Ideonella dechloratans, Pseudomonas chloritidis-mutans ASK-1.3 a group of bacteria accumulating chlorate (high chlorate accumulating perchlorate reducing bacteria, HCAP), which reduce both chlorates and perchlorate, additionally have the ability to accumulate chlorates, which can be used by the symbiotic bacteria PRB and CRB. This group includes bacteria from Dechlorosoma species HCAP-C. (4) the denitrifying bacteria belonging to the family of Rhodobacter species Haloferax denitificans and Paracoccus halodenitrificans. For these bacteria the main source of energy are nitrogen compounds, in conditions of their lack they can use perchlorate as a substrate for the reaction instead, but their importance in remediation perchlorate is low [74].

Routes of exposure

Toxins can enter the body in various ways. For perchlorate, there are potentially three possible routes of exposure: transdermal, inhalation and oral.

The transdermal route is the least important. As it is known perchlorate (in the most common chemical forms) in aqueous solutions are practically fully dissociated and ionized substances are very poorly absorbed through the skin due to epidermal barrier that effectively blocks the absorption of ions into the skin and further into the bloodstream. Therefore, it seems obvious that the potential of perchlorate entering the body through the use of contaminated water for hygiene purposes is small [37].

The inhalation route involves inhaling dust particles along with the air on which perchlorate molecules are adsorbed. In particular, perchlorate factory workers and the population of large cities where smog phenomenon occurs are the most vulnerable groups to this type of exposure [6]. Several studies [76], [77] conducted on the factory workers, who inhaled dust containing subclinical doses of perchlorate for a few hours a day showed no significant changes in the level of thyroid hormones in the blood compared to the control group, despite the increase in the content of perchlorate excreted in the urine. Study from China also informs about the increased amount of ClO₄⁻ in the urine of people exposed to inhalation contaminated with dust perchlorate. It was estimated that in the groups of toddlers, children and adults the contribution of perchlorate from indoor dust is 26, 28 and 7% of the estimated total daily intake, respectively, which makes it a significant route of exposure, especially in developmental age [78].

The oral route of exposure is the best known to date and seems to be the most important from all methods of absorbing perchlorate from the environment. Perchlorate is quickly picked up from the stomach and small intestine and then enter the bloodstream. Because of its good solubility and mobility in the aquatic environment, it can be found in many foods consumed by humans. The problem of contamination of drinking water with perchlorate was described in 1997 [1], while in recent years there has been an increasing interest in perchlorate in food products such as vegetables, fruits and even dietary supplements [9], [79]. Due to the fact that adverse effects of chronic admission of small doses of ClO₄^– for the functioning of the body are still unknown, in 2005, the US National Academy of Sciences (NAS) established a daily dose of perchlorate, which does not cause observable adverse effects at the level of 0.0007 mg/kg body weight [44], [80] and acceptable concentration of ClO₄⁻ ions in drinking waters at the level of 24.5 μg/L [81].

Absorption, distribution, metabolism, excretion (ADME)

In aqueous solution perchlorate is found in ionic form and transmitted into the blood stream from the gastrointestinal tract. Perchlorate is a poor complexing agent, therefore it is poorly accumulated in the human body. In the blood it is transported bonded with albumin, its half-life in human serum is about 6–8 h [3]. The only organs in which the perchlorate concentration to tissue concentration ratio is >1 are the thyroid (5–10x) and the skin (1–2x) [8]. When taken orally, it is absorbed very quickly and after about 10–15 min its presence in the urine can be detected. Elimination of perchlorate from the thyroid gland is relatively fast, in animal studies the half-life of ClO₄⁻ in the thyroid is about 10 h. Perchlorate is not metabolized in the human body and is also not a metabolic by-product. Some studies did not provide evidence for energetic alterations associated with perchlorate exposure at concentrations that are higher than those typically found in groundwater or surface water in the environment [82]. Thirty percent of the oral dose of perchlorate is excreted in the urine within 3 h of administration, 65–70% after 24 h, in over 95% after 48 h, hence their presence in urine is an important biomarker of exposure to ClO₄⁻ [7]. The remaining portion is removed from the body in other ways, e.g., through sweat glands and with breast milk [83].

Perchlorate in various matrices

General sources of perchlorate in environment were described in 1.1. As it is mentioned above, the most common route of exposure is via digestion system. The main source of perchlorate in human body involves contaminated water and food [16]. Numerous studies concern of perchlorate levels in ground, surface and drinking water. In the US nearly 4,000 of public water systems were surveyed and found that in 4.1% of them perchlorate concentration was above 4 μg/L. [84].

A review published in 2018 [85] has been focused on global perchlorate pollution of water. The authors concluded that most of the peer-reviewed studies (n=39; 51%) were conducted in USA and Canada and further studies in different geographical regions seem to be essential [85]. An example from India, where mean concentration of ClO₄⁻ in bottled water was 93.19 μg/L can be a confirmation of worldwide perchlorate contamination studies needs [86]. An analysis of isotopic composition of perchlorate is useful to distinguish naturally formed from man-made ClO₄⁻ [87]. Mobility of perchlorate in soils depends on many factors, such as its concentration, chemical composition and aridity of the ground etc. [88]. Although arid or semi-arid soils were investigated [8], still little is known of the presence of ClO₄⁻ in agricultural soils. Study by Calderon et al. [88] found that fertilizers, especially nitrogenous could be contaminated with perchlorate (range 12.8–95.3 mg/kg). In this way perchlorate may be absorbed to soil [88]. It is known that various plants can absorb perchlorate from soil and accumulate it in their tissues in higher concentrations than found in ground [87]. There are some studies concerning the risk of exposure to ClO₄⁻ via ingestion of fruits and vegetables. In general, the highest level of perchlorate was found in apricot from Chile (145.650 ± 4.031 μg/kg), cantaloupe form Guatemala (463.500 ± 6.364 μg/kg), spinach from USA, Mexico and Korea (39.9–187 μg/kg) [69], [89], [90]. Processed food from China was also investigated and it was found mean 6.1 μg/kg. Calculated estimated daily perchlorate ingestion for adults (EDI) of 488.5, 523.8 and 501.2 ng/kg/ bw/day in three different regions did not exceed the reference dose 700 ng/bw/day [91], but the tolerable daily intake (TDI) proposed by European Food Safety Authority (EFSA) of 300 ng/bw/day [92] was overrun. What is more, the collected data suggest that contribution of perchlorate from dietary products in relation to drinking water is near 4:1 and this ratio may be higher in children [93].

According to the scientific opinion on the risks to public health related to the presence of perchlorate in food, in particular fruits and vegetables (EFSA, CONTAM panel, 2014) [92] it is recommended to monitor the presence of perchlorate in food. The CONTAM Panel concluded that more data is needed on the occurrence of perchlorate in food in Europe, in particular for vegetables, infant formula, milk and milk products, to further reduce uncertainty in risk assessment. High levels of perchlorate have been found in the Cucurbitaceae and leafy vegetables, especially those grown under glass or sheltered [92].

Another 2017 CONTAM report confirmed that exposure to perchlorate may pose a risk to public health, so the European Commission has finally established maximum levels for perchlorate for foods containing significant amounts of perchlorate and for foodstuffs relevant to the possible risk of exposure to perchlorate. Exposure of particularly vulnerable population groups such as infants and young children [94].

On May 25, Regulation (EU) 2020/685 was published, amending Regulation (EC) No 1881/2006 concerning maximum levels for perchlorate in certain foodstuffs. In the European Union, Regulation 2020/685 applies from 1 July 2020. According to it, maximum levels for perchlorate are established for the following foodstuffs: fruit and vegetables: 0.05 mg/kg, except Cucurbitaceae and kale: 0.10 mg/kg; leafy vegetables and herbs: 0.50 mg/kg; tea (Camellia sinensis), dried: 0.75 mg/kg; herbal and fruit infusions, dried: 0.75 mg/kg; infant formulas, follow-on formulas, food for special medical purposes intended for infants and young children and infant formulas: 0.01 mg/kg; baby food: 0.02 mg/kg (young-child formulas are milk-based drinks and similar protein products intended for young children. These products are outside the scope of Regulation (EU) No 609/2013 (Report from the Commission to the European Parliament and the Council on infant formulas [COM/2016/0169 final]), and processed cereal products: 0.01 mg/kg (foodstuffs listed in this category as defined in Regulation (EU) No 609/2013 of the European Parliament and of the Council of 12 June 2013 on foods intended for infants and young children) [95]

Effects on the thyroid gland and secretion of thyroid hormones

Thyroid is the critical organ for exposure to perchlorate. The effect of perchlorate on the thyroid gland was discovered in the 1950s and 1960s, during which period these compounds were used in the treatment of hyperthyroidism [3]. Perchlorate belong to a group of compounds called goitrogens. Apart from perchlorate, goitrogens include thiocyanate (SCN⁻) and nitrate (NO₃⁻) and some other organic compounds. These compounds have the ability to inhibit the iodine uptake by the thyroid cells [96].

Iodine is necessary for the synthesis of thyroid hormones: 3,5,5′-triiodothyronine (T₃) and 3,5,3′,5′-tetraiodotyronine (thyroxine, T₄). Iodine is transported actively from the bloodstream to the follicular cells in the form of iodide ions due to the mediation of the membrane protein sodium-iodine symporter (NIS). The transport of iodides to the thyroid meets the electrochemical and metabolic criteria of active transport. It is an oxygen-dependent process which does not occur at low temperature and is also dependent on adenosine triphosphate (ATP). The first reports on the dependence of iodine transport to the thyroid in the presence of Na⁺ ions appeared in the 1950s [3]. Currently, a model in which the NIS enzyme transports two Na⁺ ions and one I⁻ ion into the cell is proposed. The presence of NIS in the tissue of the thyroid gland enables the flow of iodides from the environment, in which their concentration is very small (10⁻⁸–10⁻⁷ M), to thyroid cells, where the prevailing concentration is several hundred times greater. Such high levels of iodide are probably necessary for proper synthesis of hormones, T₃ and T_4, since patients who have been recognized to impair the ability to concentrate iodine suffer from the goiter and hypothyroidism. An important and characteristic feature of the transport of iodides to the thyroid gland is the high specificity of the enzyme for iodide ions in reference to the much higher concentrations of chloride ions. Studies have shown that the iodine ion is not the most specific anion for NIS, and a series of affinities are as follows: TcO₄⁻ ≥ ClO₄⁻ > ReO₄⁻ > SCN⁻ > BF₄⁻ > I⁻ > NO₃⁻ > Br⁻ > Cl⁻. It is noteworthy that the presented order of ions is also a lyotropic series or a series of Hofmeister [4], [97]. However, it is not entirely clear which anion property determines this series (crystal radius, hydration enthalpy, etc.). Some anions (e.g., I⁻, SCN⁻, NO₃⁻) are easily transported via NIS [98]. In case of perchlorate there is some empirical data that ClO₄⁻ may be translocated by symporter, especially in milk glands. Perchlorate was described as NIS most potent blocker, without possibility of NIS mediated transport [99]. However, latest studies shown that perchlorate inhibit iodide transport in more complex way than by the competition only. Low concentrations of ClO₄⁻ inhibit I⁻ transport also by changing the stoichiometry of I⁻ transport to 1:1, which diminishes the driving force. Electrophysiological experiments suggested that ClO₄⁻ ions may bind to a high-affinity non-transport allosteric site in NIS that prevents Na⁺ from binding to one of its two sites [11]. The harmful effect of perchlorate associated with the inhibition of iodine transport to the thyroid gland is closely related to the disturbed production of hormones T₃ and T₄ by this gland. The synthesis of these hormones is a complex process, depending on many factors. Inhibition of the transport of iodide ions to the lumen of the thyroid vesicle blocks the key stage of the synthesis of T₃ and T_4, which is the iodination of the tyrosine and its iodinated derivatives to the proper hormones. The general function of T₃ and T₄ is based on stimulation of oxygen consumption in the body cells, causing the increase of protein synthesis and obtaining a positive nitrogen balance. The result of their action is a general increase in the metabolic rate of carbohydrates and fats. They are also important modulators of developmental processes, by stimulating the pituitary gland to secrete growth hormone (somatotropin), as well as influencing the development of lungs, bones, blood vessels and CNS, which is why the group that is particularly exposed to iodine and hypothyroidism are women during pregnancy and postpartum as well as children [83], [96], [97].

Hypothyroidism manifested in fetal life or in infancy can lead to cretinism, which is characterized by low birth weight, numerous irreversible developmental defects, both physical and mental. In adults, the most common symptoms include heart failure, drowsiness, constipation, hypersensitivity to cold, brittleness of the hair and general weakness of the body [83]. Several independent recent studies suggested that various presented symptoms pertinent to autism might correlate with iodine status [100], [101].

Other toxic effects of perchlorate exposure

NIS is found not only in the tissue of the thyroid gland, its presence was also detected in eyelid body, cerebrovascular plexus, salivary and lactic glands, the latter especially during lactation. In very small amounts, NIS can be found in cellular membranes in the tissue of the heart, kidneys and lungs [99]. Unlike the thyroid gland, iodine is not metabolized in these tissues and it is reabsorbed into the circulatory system or in case of secretory glands to saliva and milk [102]. With the exception of the mammary gland, the function of iodine in these structures is poorly known and it seems that the disturbance of its transport by perchlorate does not cause changes in their physiological functioning. Doses of perchlorate used in the treatment of hyperthyroidism (400–2,000 mg/day) were fairly well tolerated by the majority of patients, there are some reports of gastrointestinal disorders, skin rashes, enlargement of the lymph nodes and fever. Studies concerned about 3–4% of patients taking small doses (400–600 mg/day), while in patients receiving higher doses (1,000–2,000 mg/day) this percentage increased to 16–18%. From severe side effects, sporadic aplastic anemia or agranulocytosis (13 people) have been reported, seven people died, and the total number of patients treated with hyperthyroidism with potassium perchlorate was greater than 1,000 [103], [104], [105], [106], [107].

No significant associations found

Pearce et al. over the years 2010–2012 published three papers [109], [110], [111] in which correlations between perchlorate and thiocyanate content and iodine concentration in urine were studied as well as the levels of thyroid hormones in pregnant women in various parts of the world. The authors measured these parameters in pregnant women in the first and second trimester of pregnancy living in Turin, Italy and Cardiff, United Kingdom. An additional criterion was the division on grounds of thyroid gland diseases - both groups were divided into normal thyroid function group and those who suffered from hypothyroidism. The median (range) concentrations were described: 98 (12–847) μg/L iodine and 2.1 (0.03–368) μg/L perchlorate in urine in hypothyroidism group and 117 (2–497) μg/L iodine and 2.6 (0.3–49) μg/perchlorate in the urine in control group in Cardiff. The median (range) concentrations in Turin were as follows: 55 (12–1129) μg/L iodine and 5.04 (0.04–108) μg/L perchlorate in urine in hypothyroidism group and 50 (6–584) μg/L iodine and 5.2 (0.2–168) μg/L perchlorate in urine in control group. Even though there was no significant correlation between perchlorate concentration in urine and thyroid hormone levels in the blood plasma, the study showed significant iodine deficiency in the studied population in Italy [109]. In next work the authors compared levels of goitrogens including perchlorate, iodine and thyroid hormones in pregnant women in the first trimester of pregnancy originating from Los Angeles, USA and Cordoba, Argentina. Perchlorate levels in urine were median (range) 7.8 (0.4–284) μg/L and 13.5 (1–676) μg/L in Los Angeles and Cordoba, respectively, and iodine levels were found 144 (16–733) μg/L and 130 (9–693) μg/L in Los Angeles and Cordoba, respectively. No statistically significant relationship was found between the concentration of perchlorate in the urine and the level of thyroid hormones in the blood, however, attention was paid to a significantly higher (p<0.003) perchlorate level in women in Argentina [110]. The third study was carried out in Athens, Greece, where women in the first trimester of pregnancy were examined analogously to the previous ones in order to determine the concentrations of perchlorate, iodine and thyroid hormones. It was found 4.1 (0.2–118.5) μg/L median (range) of perchlorate and 120 (28–538) μg/L of iodine in the urine, respectively. With respect to the entire studied population, a significant relationship was found between free T₃ (FT₃) and free T₄ (FT₄) levels and urinary perchlorate concentration, yet in the subgroup of women with iodine concentration <100 μg/L in the urine this relationship was not demonstrated [111]. All perchlorate determinations were performed using the IC-ESI-MS technique described elsewhere [112]. The measurement sensitivity was sufficient, the detection limit was 0.003 μg/L, and the measurement range of concentration was 0.025–100 μg/L. The authors of the above studies concluded that perchlorate have no significant effect on thyroid functions when exposed to environmental conditions. Importantly, there is a relatively small number of participants in each study. Note that the largest number has been studied in Pearce et al. [109], the smallest in Pearce et al. [111] (see Table 4). Additionally, there is a lack of estimated daily dose of perchlorate taken by the examined patients. No statistically significant relationship between perchlorate content in urine and thyroid hormone levels in pregnant women was also found in the paper by Leung et al. [113], where the concentration of perchlorate in breast milk was studied as well. The following perchlorate concentration results were obtained in milk, urine of mothers and urine of infants: 4.4 (0.5–29.5), 3.1 (0.2–22.4) and 4.7 (0.3–25.3) μg/L median (range), respectively. Concentrations of iodine in milk, in urine of mothers and in urine of infants was as follows 45.6 (4.3–1,080), 101 (27–570) and 197.5 (40–785) μg/L median (range), respectively. However, these studies were based on a population from one selected area (Boston, USA), a small sample size (N=64 mother-child pairs) and a socioeconomic group declaring an appropriate diet, especially in the context of sufficient iodine intake. It seems, therefore, that the results obtained in this study cannot be unambiguously extrapolated to estimate the effect of perchlorate on the level of thyroid hormones in breastfed babies. Mortensen et al. [114] investigated the effect of perchlorate, thiocyanate and nitrate on the plasma thyroid hormone concentration in women in the third trimester of pregnancy. The study group consisted of 359 women from various locations in the United States. The results of concentrations 4.03 (3.72–4.36) μg/L geometric mean, (95% CI) of perchlorate and median (range) 167 (107–217) μg/L of iodine in urine were obtained, with no significant relationship between perchlorate concentration in urine and the level of TSH and FT₄; however, iodine deficiency in the study population has not been found as well. Gold et al. [115] examined the impact of environmental exposure to perchlorate in the population with various levels of exposure, and there was no correlation between the level of exposure and thyroid function, however, data on the concentration of perchlorate and iodine in urine were not presented by the authors.

Significant associations found

On the other hand, there are several papers that also described the effect of perchlorate concentrations in the body fluids on the level of thyroid hormones in the blood, indicating the existence of statistically significant relationships between these parameters. Most of the reviewed studies concerned to determine perchlorate in urine, although there is an interesting work that describes a positive relationship between the concentration of ClO₄⁻ in the colostrum and the level of TSH in mothers, nonetheless without indicating the significant correlation in the case of TSH in their children [116]. The results described in this paper cover only 185 mother-child pairs, so it seems to be insufficient to be representative of the general population.

Three reviewed studies were based on data collected during NHANES 2007–2008. Paper by Mendez and Eftim [117] uses specially designed additive models (GAMMs) to estimate the effect of perchlorate, phthalates and other NIS inhibitors on the level of thyroid hormones. In a sample of 970 men and 907 women, a significant negative correlation was found between the concentration of perchlorate in the urine and the concentration of triiodothyronine and thyroxine in the plasma. Another work [118] which is based on data obtained from NHANES 2007–2008 describes the effect of environmental exposure to perchlorate on thyroid hormone levels. The results indicate that people classified as high-risk exposures (1,939 people) had an average of 5% lower plasma concentrations of thyroxine than those from the low-risk group (2,084 people). Moreover, people with a high content of perchlorate, thiocyanate and low iodine concentration in the urine (n=62) had thyroxine at 12.9% lower (mean difference = 1.07 mg/dL, 95% CI=0.55–1.19) than in patients with low perchlorate, thiocyanate, and normal urine iodine (n=376). Undoubtedly, a great advantage of this study appears to be a numerous test group as well consideration of other factors i.e., thiocyanate and iodine, apart from perchlorate. Note that the research roughly confirms previous reports about the negative impact of NIS inhibitors on thyroid function.

Drawing on NHANES 2007–2008 Przybyla et al. [119] examined the relationship between the content of 11 different compounds (perchlorate, some phenolic compounds and phthalates) in the urine and the level of thyroid hormones. The study group consisted of 850 men and 710 women, with an mean concentrations 3.48 and 4.37 μg/L of perchlorate in the urine, respectively. What is interesting is that authors describe the negative impact on thyroid function (a significant negative relationship with the level of T₄ in the plasma), when the cumulative effect of all tested compounds is taken into account. Additionally, it was observed only for men. The possible explanation given by the authors is the disruption of liver function, observed in animal studies, probably caused by mentioned goitrogens, which can reduce the conversion of T₃ to T₄. For the perchlorate itself, the statistically significant relationship was not confirmed.

Studies based on NHANES results have the advantage of a very large study group, but they provide results only for US population. There were a number of studies, where the authors examined some specific populations or they were focused on different than previously described objectives. Charatcharoenwitthaya et al. [120] determined these parameters in 200 women in the first trimester of pregnancy from Thailand, obtaining median (range) 1.9 (0.1–35.5) μg/L of perchlorate and 153.5 (50–1219) μg/L of iodine in the urine. There was a relevant positive correlation between TSH concentration and negative relationship between FT₄ level in plasma and urine perchlorate content. What is more, previous studies by the authors [109], [110], [111] found no significant correlations. The differences between mentioned studies might be caused by higher risk of exposure on perchlorate in Thai group, especially via foodstuffs and drinking water. Horton et al. [121] have examined a 284 women in the first trimester of pregnancy in the area of New York, USA. The mean ± standard error concentrations of perchlorate and iodine in urine were 3.54 ± 0.02 mg/g creatinine and 235.39 ± 39.40 mg/g creatinine, respectively. A statistically significant relationship was found between the level of TSH in plasma and the concentration of perchlorate in urine calculated as creatinine. The undoubted advantage of this study is applying a novel statistical methodology (WQS) to assess combined effects of more than one thyroid disrupting factor. But relatively inconsiderable sample size and small study area can lead to the risk of bias. The studies on much larger samples than mentioned above were carried out by Steinmaus et al. in 2016 [122]. Perchlorate and iodine in the urine and thyroid hormone level in blood plasma were determined in 1880 women in the first trimester of pregnancy in the area of California, USA. The median (range) level of 6.50 (0.23–177) µg/L perchlorate and 154.5 (0.8–3,000) µg/L of iodine in urine were found. It is interesting to note that the ClO₄⁻ concentration in the urine was almost twice higher than in the general population of the US. A significant positive correlation was found between the logarithmic values of urine perchlorate concentration and TSH in the serum and the negative correlation between the logarithmic T₄ and FT₄ values in serum and urinary perchlorate. The obtained dependences are in line with previous data obtained during NHANES 2002–2003. An unexpectedly strong correlation between the concentration of perchlorate (as well as thiocyanate and nitrate) in the urine and the level of T₄ on a sample of 92 children aged 1–12 months was described by Cao et al. [123]. In addition, perchlorate and TSH were positively related. Also, Knight et al. [124] confirms the critical negative correlation between urinary perchlorate concentration and serum FT₄ level. What is more, there was no correlation between the level of ClO₄⁻ and iodine in urine. The study was conducted in the South West of England on a group of 308 women in the third trimester of pregnancy.