Guidelines for Canadian Drinking Water Quality: Guideline Technical Document – Antimony

Antimony may enter the aquatic environment by way of natural weathering of rocks, runoff from soils, effluents from mining and manufacturing operations, and industrial and municipal leachate discharges.^{Footnote 4Footnote 5} Household piping and possibly non-leaded solders are sources of antimony in tap water, as soft water may leach antimony from the pipes. The potential for antimony to leach from copper pipes with solder joints made from 95% tin/5% antimony into high-purity, tap and well water samples has been investigated.^{Footnote 6} The high-purity water had a pH of 6.8; the pH, alkalinity and chloride values of the tap water were 7.8, 30 mg/L (as CaCO₃) and 4 mg/L, respectively, whereas the corresponding values for the well water were 8.1, 155 mg/L and 19 mg/L. There was no detectable leaching (detection limit 1.2 µg/L) of antimony into the high-purity water or the well water; however, leaching occurred into tap water, but only after 7 days of contact. After 7, 28 and 90 days' contact time, the amount of antimony leached into tap water resulted in final concentrations of 2.0, 3.7 and 7.3 µg/L, respectively. It is noted that no tests were conducted with acidic (below pH 6) waters.

Antimony is released into the air as stack dust from industrial sources, such as coal-fired power plants, inorganic chemical plants and metal smelters.^{Footnote 7}

Antimony is not degradable per se. The two main factors that influence the behaviour of antimony are its solubility and complexation. Antimony compounds range in solubility from insoluble to fully soluble. Most inorganic antimony compounds are insoluble, whereas compounds that become attached to organic ligands are soluble.^{Footnote 8} Some antimony salts, such as the trichloride, sulphate, potassium tartrate and pentachloride, are soluble in water, whereas the oxides tend to precipitate.^{Footnote 9} Sorption may also be a key environmental process, but the available information is incomplete. Antimony does not bioaccumulate.

The fate of antimony in the aquatic environment is determined by several factors, including pH, oxidation-reduction potential, sorptive interactions and biologically mediated methylation. In the aquatic environment, antimony can be present as an ion or a soluble complex. Most of the dissolved antimony (in the +5 oxidation state)^{Footnote 10} that might be discharged to natural water would soon precipitate as antimony trioxide or antimony pentoxide and be removed by sedimentation.^{Footnote 9}

Exposure

Canadians can be exposed to antimony via water, air, food and urban dust. The concentrations of antimony detected in Canadian surface water range from 0.001 to 9.1 mg/L.^{Footnote 11} Concentrations are typically less than 10 µg/L and are often closer to 1 µg/L.^{Footnote 12} In a recent survey of water supplies in Newfoundland,^{Footnote 13} samples were taken from both homes and raw water sources (surface water and groundwater); antimony was found at concentrations below 5 µg/L, regardless of the source or type of water. The levels of antimony in community drinking water supplies in New Brunswick were generally less than 5 µg/L; however, in one well water sample, antimony was reported at 17 µg/L.^{Footnote 14} In a survey of municipal water supplies of 88 cities in the United States, the levels of antimony found in tap water were on average less than 0.2 µg/L.^{Footnote 15} In an analysis of 3834 drinking water samples collected from randomly selected households in 35 geographic areas of the United States, antimony was found in 16.5% of the samples, at concentrations ranging from 0.6 to 4.0 µg/L, with a mean value of 1.87 µg/L.^{Footnote 16} The average intake of antimony from drinking water has been estimated to be less than 8 µg/d.^{Footnote 17}

Antimony concentrations are higher in urban air than in non-urban air, presumably due to fossil fuel combustion, automobile emissions and incineration in urban areas. The mean, median and maximum concentrations of antimony in aerosols at three sites in Quebec, Ontario and Nova Scotia were 0.04-0.10, 0.11-0.23 and 0.37-2.17 ng/m³, respectively.^{Footnote 18} Twenty-four-hour samples collected at 10 locations in Washington, DC, yielded average antimony concentrations ranging from 1.1 to 3.0 ng/m³.^{Footnote 19}

Concentrations of antimony in meat, freshwater fish, poultry, cereals, fruit and vegetables appear to range from about 1 to 10 ng/g wet weight; significantly higher levels have been measured in marine organisms.^{Footnote 7} The daily intake of antimony from food has not been estimated for Canadians. A study was conducted in 1967 to measure the concentrations of antimony in the diets of children (aged 9 to 12) in 28 U.S. cities. Concentrations of antimony (measured by atomic absorption spectrophotometry) ranged from 0.209 to 0.500 mg/kg of food (weighted average 0.361 mg/kg), resulting in daily dietary intakes between 0.25 and 1.28 mg.^{Footnote 20} These values are thought to be too high because of the analytical method used.^{Footnote 21} Iyengar et al.^{Footnote 22} estimated the intake of antimony from food for 25- to 30-year-old males in the United States by analysing 201 foods (by instrumental neutron activation analysis) from the U.S. Food and Drug Administration's Total Diet Study. Antimony was found at a mean concentration of 0.0093 ± 0.0014 µg/g dry weight (mixed-diet composite); assuming an adult male consumes approximately 3075 g wet weight or 498 g dry weight (16.2% of wet weight), the authors suggested a total antimony intake of approximately 4.6 µg/d from food.

The information on levels of antimony in Canadian drinking water, air and food was insufficient to allow the determination of national intakes by these three routes of exposure. U.S. exposure levels and intakes are probably a reasonable approximation of Canadian exposure levels and intakes because of the similar environment and food habits in North America. Assuming a contribution of 4.6 µg from food (as calculated above),^{Footnote 20}0.04 µg from air (calculated using the maximum measured Canadian concentration of 2.17 ng/m³and a daily air intake of 20 m³)^{Footnote 18} and 2.8 µg from water (calculated from the mean value of 1.87 µg/L and an average daily water consumption of 1.5 L),^{Footnote 16} an adult would receive 7.44 µg of antimony per day; approximately 38% of this total intake would come from drinking water.

Analytical Methods and Treatment Technology

Antimony may be analysed in water using an atomic absorption spectrophotometer equipped with a graphite furnace.^{Footnote 6} The detection limit given by the laboratory was 1.2 µg/L (based on three times the standard deviation of the blank). The practical quantitation level, based on a minimum of five times the detection limit, is 6 µg/L. Although more sensitive methods are available for the analysis of antimony in water,^{Footnote 23} lower detection limits are not currently readily achievable by many monitoring laboratories.

Little information is available on the removal of antimony from source water during water treatment. However, as antimony is chemically similar to arsenic, the methods used for the removal of arsenic (which can be removed to <5 µg/L from groundwater containing natural arsenic levels as high as 1.0 mg/L using activated alumina columns)^{Footnote 24Footnote 25} could possibly be used for antimony removal.^{Footnote 26} Coagulation with alum or ferric salts, particularly in the presence of turbidity, may have potential for conventional treatment. Lime softening can very effectively remove antimony from hard waters. The removal will be dependent on the pH and antimony valence. In bench-scale tests, a water-insoluble starch xanthate, which acts as an ion exchange material, has been reported to be effective in removing antimony from wastewater (from 5 mg/L to roughly 0.01 mg/L).^{Footnote 8}

Because antimony may also be introduced to drinking water after the water leaves the treatment plant as a result of leaching from materials in the distribution system or household plumbing, corrosion control is a more effective method of preventing high concentrations of antimony at the point of consumption. Adjustment of the pH from less than 7 to 8 or 9 and moderate increases in alkalinity will reduce the corrosiveness of the water and minimize leaching. Corrosion inhibitors may also be added to the water.^{Footnote 6}

Health Effects

Pharmacokinetics

Absorption

The respiratory tract is the principal route by which antimony enters the body in occupational exposure. Precise information regarding respiratory absorption is lacking,^{Footnote 21} although it is known to be correlated with particle size.^{Footnote 27} Absorption of antimony from the gastrointestinal tract is thought to occur slowly and in small amounts.^{Footnote 28} The extent of gut absorption depends on solubility and chemical form. After their administration (as 1-mL solutions of ¹²⁴Sb tartrate) by gavage to eight Syrian hamsters, only 1.6 and 2% of the initial body burdens of trivalent and pentavalent antimony, respectively, were present on day 4, 61% and 64% of which were found in the gastrointestinal tract.^{Footnote 29} The median value of 1.6% includes two hamsters that received the largest amount of trivalent antimony, which retained 15% and 9% of the initial body burden on day 4.

About 15% of a single oral dose of soluble isotopic antimony potassium tartrate (4.4 mg/kg bw, or 7 µg of antimony) was absorbed in the gastrointestinal tract of white rats; absorption was determined following recovery of antimony in urine, faeces and tissues.^{Footnote 30} Cows administered a single oral dose of ¹²⁴Sb trichlorides (21 mg Sb/kg bw) absorbed less than 5% in their gastrointestinal tract.^{Footnote 31} No estimate of gastrointestinal absorption in humans has been reported.

Distribution

The distribution of antimony within the body and its excretion are a function of both the route of administration and the valence state of antimony. After inhalation, tissue distribution studies show that the trivalent form accumulates more rapidly than the pentavalent form in the liver, whereas pentavalent antimony is found preferentially in the skeleton. In the blood, the trivalent form is lodged primarily in red blood cells, whereas the pentavalent form is carried by plasma.^{Footnote 29} Following ingestion in animals, liver, kidney, bone, lung, spleen and thyroid are the major sites of accumulation outside the gastrointestinal tract.^{Footnote 10} In humans, the therapeutic use of trivalent antimony against parasites leads to higher accumulation in the liver, the thyroid and the heart.^{Footnote 32}

Antimony was found in placental tissues, milk, amniotic fluid and the umbilical cord of pregnant or nursing women who had worked in smelters.^{Footnote 33} In mice, antimony appears to cross the placental barrier more readily after intraperitoneal injection than when given in the diet. The newborn mouse is also able to take up antimony from its mother's milk when the mother is fed a contaminated diet during pregnancy and after delivery.^{Footnote 34} Antimony concentrations in breast milk were measured in 21 Italian women (>130 samples) during a 2- to 3-month period starting 15 days after childbirth. Concentrations ranged from less than 0.05 to 12.9 ng/g, with a mean of 3.0 ± 0.4 ng/g.^{Footnote 35}

Excretion

Excretion rates and routes differ depending on animal species, routes of administration and the valence of antimony. In general, trivalent antimony is excreted primarily in faeces, whereas pentavalent organic antimony is excreted primarily in urine.^{Footnote 36} In a study performed to investigate the metabolism of antimony,^{Footnote 31} three cows (in a lactating phase) were given antimony (as ¹²⁴SbCl₃) orally at 2.84, 2.72 and 2.00 mCi, respectively (corresponding to an average dose of 21.1 mg Sb/kg bw). The excretion of antimony in faeces accounted for 82% of the dose, whereas milk and urine accounted for 0.008 and 1.1%, respectively. The spleen, liver, bone and skin contained the highest tissue concentrations. In a parallel study performed by the same authors, one cow received an intravenous injection of ^124SbCl3 ^{(0.234 mCi), which} corresponded to a dose of 1.5 mg Sb/kg bw. Faecal excretion accounted for 2.4% of the administered dose, whereas 0.08% and 51% of the total dose were excreted in milk and urine, respectively.

In humans, urine accounts for 1.2-3.6 µg of daily antimony excretion. Approximately 0.3-0.9 µg/d is excreted in faeces, and less than 1 µg/d is excreted by other routes.^{Footnote 8}

Effects in Humans

Acute and Subchronic Toxicity

Antimony poisoning has resulted from accidental occupational inhalation, ingestion of food contaminated by storage containers and therapeutic treatment with tartar emetic (potassium antimony tartrate).^{Footnote 37} Antimony compounds have been used for a long time as therapeutic agents for parasitic diseases such as schistosomiasis, leishmaniasis, trypanosomiasis and ulcerative granu-loma. Side effects of antimony therapy (the average dose is up to 1 g/d for 10 days) include myocarditis, hepatitis and nephritis.^{Footnote 4}

Acute intoxication is characterized by abdominal pain, vomiting, diarrhoea, dehydration, muscular pain, shock, haemoglobinuria, anuria and uraemia.^{Footnote 37} In addition, severe myocardial symptoms and convulsions have been observed with acute doses of antimonials, and some deaths were attributed to liver necrosis. In humans, the lowest lethal dose by ingestion was 0.75 mg potassium antimony tartrate/kg bw.^{Footnote 38} Antimony levels of about 30 mg/L in a contaminated drink caused nausea, vomiting and diarrhoea in 150 children.^{Footnote 39}

Dermal contact with antimony can cause eczema and dermatitis, resulting in papular, vesicular and pustular eruptions.^{Footnote 37} Symptoms of gastrointestinal disturbance and mild jaundice appear 2-3 weeks after a course of injections of antimony compounds.^{Footnote 40}

Chronic Toxicity and Carcinogenicity

Chronic exposure to lower doses of antimony compounds is primarily associated with myocardial effects. Heart complications, altered electrocardiogram (ECG) patterns and sudden deaths were observed in eight out of 125 workers who were occupationally exposed to antimony trisulphide at concentrations ranging from 0.58 to 5.5 mg/m³ (average 3.0 mg/m³) for 8-24 months.^{Footnote 41} Seven out of 111 (6.3%) surviving workers were found to have ulcers, compared with 59 out of 3912 (1.5%) control group workers.

In a recent mortality study, the lung cancer death rates of 1014 men (91.5% of Spanish ancestry) employed between 1937 and 1971 in an antimony smelter in Texas were compared with ethnic-specific lung cancer death rates in Texas. Mortality from lung cancer among antimony workers was elevated (standardized mortality ratio [SMR] 1.39; 90% confidence interval [CI] 1.01-1.88), and a significant positive trend in mortality with increasing duration of employment was observed. A significant increase in mortality from cancers of the liver, biliary tract and gall-bladder (SMR 3.17; 95% CI 1.27-6.52) was observed in the smelter workers compared with white males.^{Footnote 42} In a study of men employed in a British antimony smelter between 1961 and 1992, a significant increase in deaths from lung cancer was found in the census p opulation of men working at the smelter before 1961, but not in the prospective cohort recruited after that date. It was not possible to attribute the increased risk of lung cancer to any particular agent, as the antimony workers were exposed to a variety of agents in addition to antimony and its oxides, including arsenic, sulphur dioxide and polycyclic aromatic hydrocarbons. No relationship was found between length of time worked and mortality from lung cancer.^{Footnote 43}

Reproductive Toxicity

A higher incidence of menstrual disorders and spontaneous late abortions (77.5% and 12.5%, respectively) was reported in female smelter workers who were exposed to metallic dust and antimony trioxide and pentoxide compared with a control group (56% and 4.1%, respectively).^{Footnote 33} The weight of the babies born to these exposed women was slightly less at three months of age and was significantly less at one year of age. Antimony was also found in breast milk (3.3 ± 2 mg/L), placental tissue (3.2-12.6 mg/100 mg) and amniotic fluids (6.2 ± 2.8 mg/100 mg).

Effects in Laboratory Animals and In Vitro Test Systems

Acute and Subchronic Toxicity

The acute oral LD₅₀ value for potassium antimony tartrate in mice is 600 mg Sb/kg bw.^{Footnote 44} Minimum LD₅₀s for antimony of 100 and 150 mg/kg bw have been reported in rats and in guinea-pigs, respectively, after intraperitoneal injection; death was attributed to myocardial failure. In contrast, the minimum LD₅₀s for potassium antimony tartrate were 11 and 15 mg Sb/kg bw after intraperitoneal injection. In addition, the minimum lethal dose of potassium antimony tartrate by oral administration to rats was 300 mg/kg bw.^{Footnote 45} The subcutaneous and intravenous LD₅₀values in mice were 20 and 24 mg/kg bw, respectively.^{Footnote 46} Oelkers^{Footnote 47} reported that a single oral dose of 115 mg potassium antimony tartrate/ kg bw was fatal to 50% of rabbits. A single oral dose of potassium antimony tartrate administered to rats at 300 mg Sb/kg bw resulted in death, attributed to myo-cardial failure.^{Footnote 45} However, no rats died following a single exposure to 188-16 714 mg Sb/kg bw or lower as inorganic antimony or to a dose of 7000 mg antimony as metallic antimony.

An increase in non-protein nitrogen in the blood and urine was reported in rabbits orally fed a dose of 15 mg potassium antimony tartrate/kg bw in a subchronic study.^{Footnote 48} Jaundice was noticed during the last two days of exposure, and the livers of some of the animals showed fatty degeneration and parenchymal necrosis. Following administration of 25 subcutaneous doses of antimony tri-fluoride (15 mg/kg bw) in rats over a one-month period, the rats developed oedema, fatty infiltration and cloudy swelling in the liver; in the kidney, swelling of epithelial cells lining the convoluted tubules, desquamation of the epithelium, protein masses in tubular lumina and shrunken glomeruli were observed. However, in a second group injected subcutaneously with antimony trioxide (at an equivalent dose of antimony), few pathological changes of significance were found, suggesting that at least some of the changes observed with antimony tri-fluoride were attributable to the presence of fluorine.^{Footnote 49} Severe diarrhoea was observed in dogs administered antimony trioxide at a dose of 84 mg Sb/kg bw per day for 32 days. No effects were observed in rats exposed to 501 mg Sb/kg bw per day or less as antimony trioxide for 20 days.^{Footnote 50}

Rats, rabbits and dogs were exposed to the dust of antimony trisulphide for seven hours per day, five days per week, for at least six weeks, at concentrations of 3.07, 5.6 and 5.32 mg/m³, respectively.^{Footnote 41} Rats and rabbits developed parenchymatous degeneration of the myocardium and functional disorders of the heart, accompanied by ECG alterations (e.g., flattened T-wave pattern); these findings were not as pronounced in dogs.

In a range-finding test for "subacute" oral toxicity,^{Footnote 51} rats (five per sex per group) were fed a diet containing antimony trioxide at doses ranging from 60 to 1070 mg/ kg bw per day for 30 days. The animals showed reduced growth and appetite as well as unspecified micropatho-logical changes in the liver, kidney, spleen or testes at the highest dose. No effects were observed at or below 270 mg/kg bw per day.

Growth retardation as a result of subchronic oral exposure via feeding has also been suggested by a decrease in body weight gain in a study in which antimony or antimony trioxide was administered in the diet (0.1% or 1.0% antimony, 1.0% antimony trioxide) of 10 male Wistar rats for 12 weeks. The average weights of the rats at the end of the experiment were 438.0 ± 22.3 g (0.1% antimony), 401.0 ± 18.3 g (1.0% antimony) and 454.3 ± 38.8 g (1.0% antimony trioxide), compared with 489.5 ± 60.5 g for rats in the control group. Body weights returned to normal after 12 weeks on an antimony-free diet.^{Footnote 52}

In a recently completed study,^{Footnote 53} male and female Sprague-Dawley rats (15 per sex per group) were exposed to a soluble trivalent antimony salt (potassium antimony tartrate) in tap water at concentrations of 0.5, 5, 50 or 500 mg/L for 13 weeks. Calculated antimony intakes ranged from 0.06 to 42.17 mg/kg bw per day in males and from 0.06 to 45.69 mg/kg bw per day in females (although the actual "absorbed" doses were closer to 0.006-4.5 mg/kg bw per day based on the estimation of a 10% gastrointestinal absorption of potassium antimony tartrate). Control rats received unsupplemented tap water as drinking water. An additional 10 male and 10 female rats were included in each of the control and 500 mg/L groups and were given tap water for a further four weeks following the 13-week exposure period. All animals survived the treatment and recovery periods with no clinical signs of toxicity. Rats exposed at the highest dose level consumed significantly (35%) less water than rats in the control and lower dose groups and showed suppressed body weight gain. During recovery, water intakes were quickly restored to that of the control groups, and body weight gain was accelerated. At necropsy, one male in the 5 mg/L dose group and three males in the highest dose group had gross haematuria. A male rat in the highest dose group had a cirrhotic liver, and a female rat in the lowest dose group had a nodular, fibrotic spleen; because this effect was not observed at higher doses, the authors considered it to be not biologically significant.^{Footnote 54} Females in the 50 mg/L dose groups had a significant decrease in thymus to body weight ratios compared with controls. Significantly increased kidney to body weight ratios were observed in the highest dose group of both males and females compared with the control males and females. Both males and females in the highest dose group had decreased alkaline phos-phatase activity and creatinine serum levels, whereas only males at 500 mg/L had decreased red blood cell and platelet counts and a slight increase in mean corpuscular volume. Females showed a dose-related decrease in serum glucose starting at 5 mg/L. Mild adaptive histological changes were observed in the highest dose groups in the thyroid, liver and pituitary glands of both sexes and in the spleen of male rats and thymus of female rats. After a four-week recovery period, the pituitary gland appeared largely normal, and the changes in the liver and thyroid of both sexes became less severe. On the other hand, minimal changes persisted in the spleen of both sexes and in the thymus of males. Tissue antimony levels were dose-related in the following (descending) order: red blood cells >> spleen, liver > kidney > brain, fat > serum. Following the recovery period, the antimony level in the highest-dose animals decreased in all tissues (antimony levels in red blood cells were not determined) except the spleen. A no-observed-adverse-effect level (NOAEL) of 0.5 mg antimony/L in drinking water, equivalent to an a verage intake of 0.06 mg/kg bw per day, was established based on the histological changes observed at 5 mg/L.

Decreases in haematocrit, haemoglobin and plasma protein levels were observed in rats exposed to 500-1000 mg Sb/d for 12-24 weeks.^{Footnote 52Footnote 55} Sunagawa^{Footnote 55} observed a decrease in red blood cell count in rats exposed to 418 mg Sb/kg bw per day as antimony trioxide for 24 weeks. Cloudy swelling of the hepatic cords was also observed at a dose of 418 mg Sb/kg bw per day as antimony trioxide or to 500 mg Sb/kg bw per day as metallic antimony. However, lower concentrations of antimony trioxide or potassium antimony tartrate did not produce the effects mentioned above.^{Footnote 50Footnote 56}

When dogs were treated with 6644 mg Sb/kg bw per day for 32 days, weakness and difficulty in moving hind limbs were observed.^{Footnote 50}

Chronic Toxicity and Carcinogenicity

There are only two studies available on the chronic toxicity of antimony.^{Footnote 56Footnote 57} In the first,^{Footnote 56} potassium antimony tartrate was provided in drinking water (5 mg/L as antimony) of Charles River CD strain mice (54 per sex per group) from the time of weaning until natural death. The animals were weighed weekly up to eight weeks and then monthly. The daily intake of antimony was calculated by the authors to be 350 µg/kg bw. Antimony did not significantly suppress growth in either male or female mice during the first year but did result in weight loss in males after 18 months (p < 0.025) and decreased weight gain in females measured at 12 and 18 months (p < 0.005). The median life spans and 75% life spans of female mice were shortened by 49 and 86 days, respectively. Effects on the life span were minimal in males.

In the second study,^{Footnote 57} potassium antimony tartrate was provided in drinking water (5 mg/L as antimony) to Long-Evans rats (50 per sex per group) from the time of weaning until natural death. Water consumption by the rats was not reported. However, water consumption for rats of the same strain handled in the same laboratory was reported as 7.5 mL/100 g bw per day for females and 6.8 mL/100 g bw per day for males.^{Footnote 58} These would correspond to estimated doses of 340 µg/kg bw per day and 375 µg/kg bw per day for male and female rats, respectively. Antimony was more toxic to rats than to the mice in the first study. The life spans of the exposed rats were shortened significantly (the median life span was 106-107 days less than controls, and the 90% life span was 70 and 165 days less than controls for males and females, respectively). Longevity, the mean age of the last surviving 10%, was also significantly reduced compared with control rats. There were no measurable differences in the body weights between exposed and control rats. However, antimony accumulated in soft tissues with age. The mean concentration of antimony in five tissues (kidney, liver, heart, lung and spleen) of all rats analysed was 13.1 µg/g. As antimony was innately toxic to rats at 5 mg/L (which corresponded to 0.340 and 0.375 mg Sb/kg bw per day for males and females, respectively), this level was a lowest-observed-adverse-effect level (LOAEL) in this study.

The same two oral lifetime studies in mice^{Footnote 56} and rats^{Footnote 57} have suggested that antimony is not a carcinogen. In the mouse study,^{Footnote 56} dead animals were necropsied, gross lesions were recorded and tissues were examined histologically. Tumours identified at necropsy were also examined microscopically. Tumours were found in 34.8% of control animals and 18.8% of antimony-exposed animals. No reason was given for the high tumour incidence in controls. In the rat study,^{Footnote 57} antimony was not tumorigenic, as evidenced by a comparison of visible tumours in controls and treated animals at necropsy. Tumour frequencies in male and female control rats were 20.0% and 35.9%, respectively, whereas frequencies in treated male and female rats were 12.0% and 38.3%, respectively. In a third study, administration of 5 mg Sb/L (as the potassium tartrate) in drinking water to Charles River CD-1 strain mice for life did not result in any significant differences in the incidence of spontaneous or malignant tumours.^{Footnote 58} It should be noted that these studies are old and inadequate by modern standards. Major deficiencies of all three studies included the lack of complete histology, administration of only one dose level and limited reporting of pathology.

The induction of lung neoplasms was reported in female CDF rats, but not in male CDF rats, following the inhalation of antimony trioxide or antimony ore concentrate (concentrations ranged from 1.6 to 45 mg/m³ in the two studies) for periods of up to a year, followed by surveillance from 20 weeks to 15 months.^{Footnote 59Footnote 60}

Mutagenicity

Several antimony compounds showed mutagenic potential in short-term in vitro tests. They have caused chromosome breaks in cultured human leukocytes^{Footnote 61} and enhanced the transformation of hamster cells by SA virus 7.^{Footnote 62} Three antimony compounds (antimony trioxide, antimony trichlorides and antimony pentachloride) were positive in the Bacillus subtilis rec assay, inhibiting the cellular growth of a recombinant-deficient strain of B. subtilis more strongly than that of a wild one.^{Footnote 63}

Reproductive and Developmental Toxicity

The information gathered from very limited animal studies on reproductive and developmental effects suggests that prenatal exposures to antimony can lower the rate of conception.^{Footnote 33} Female rats were exposed either to antimony dust via a single intraperitoneal injection of 50 mg/kg bw (acute exposure) or to antimony trioxide dust for four hours per day for 1.5-2 months at a concentration of 250 mg/m³ (subchronic exposure).^{Footnote 33} The females were mated 3-5 days following injection, whereas inhalation exposure was continued throughout the gestation period. Following mating, only 15 of 30 acutely exposed females and 16 of 24 subchronically exposed females conceived, compared with nine of 10 females and 10 of 10 females in the two control groups. Fewer offspring were born in both exposed groups (average 6.2 per litter) than in the control group (8.3 per litter). No morphological changes were observed in the foetuses. Resorption and foetal deaths were not discussed.

In a study to evaluate the effect of antimony on development of vasomotor reactivity in the pups,^{Footnote 64} female albino rats (30 per group) were given antimony trichlorides at 0, 0.1 or 1 mg/100 mL (0, 1 or 10 mg/L) in drinking water ad libitum from the first day of pregnancy until weaning of the pups (22 days after delivery). The pups (10 per dam) received antimony at 0, 1 or 10 mg/L in their drinking water ad libitum from weaning to day 60. The dams showed a significant dose-related decrease (p < 0.05) in body weight on gestational day 20, but not on day 10. The pups in the high-dose group also had significantly reduced body weight (p < 0.05) from the 10th to the 60th day of age. Because food and water consumption data were not provided, it is not possible to determine if the effects on body weight were due to the direct toxicity of the chemical or to reduced food and/or water intake. The response of pups to hypertension- and hypotension-inducing drugs was altered at both antimony doses.

No foetal abnormalities were observed on gestational day 20 in pregnant rats after intramuscular administration of a solution containing antimony dextran glycoside (125 or 250 mg Sb/kg bw) on five occasions between the eighth and 14th days of gestation. Antimony was not detected in the foetuses.^{Footnote 65}

Classification and Assessment

Antimony is classified as having inadequate data for the evaluation of its carcinogenicity (Group V). For compounds classified in this way, the tolerable daily intake (TDI) is derived on the basis of division of the NOAEL or LOAEL in an animal study by an appropriate uncertainty factor.

Based on the available information, it is clear that antimony affects many organs and systems in the body. In a recently completed 13-week study using the most appropriate route and vehicle of administration (i.e., drinking water), groups of rats were exposed to potassium antimony tartrate in drinking water at concentrations of 0.5, 5, 50 or 500 mg/L.^{Footnote 53} A NOAEL of 0.5 mg Sb/L in drinking water, equivalent to an average intake of 0.06 mg/kg bw per day, was established based on the histological changes observed at 5 mg/L. For antimony, the TDI is derived as follows:

$\text{TDI}=\genfrac{}{}{0.1ex}{}{0.06\text{mg/kg bw per day}}{300}=0.0002\text{mg/kg bw per day}$

where:

• 0.06 mg/kg bw per day is the NOAEL in a 13-week study in rats^{Footnote 53}
• 300 is the uncertainty factor (×10 for intraspecies variation; ×10 for interspecies variation; and ×3 for the use of a short-term study).

Rationale

The maximum acceptable concentration (MAC) for antimony in drinking water is derived from the TDI as follows:

$\text{MAC}=\genfrac{}{}{0.1ex}{}{0.0002\text{mg/kg bw per day}\times 70\text{kg}\times 0.38}{1.5\text{L/d}}\approx 0.004\text{mg/L}$

where:

• 0.0002 mg/kg bw per day is the TDI, as derived above
• 70 kg is the average body weight of an adult
• 0.38 is the proportion of total daily intake allocated to drinking water (as calculated in the "Exposure" section)
• 1.5 L/d is the average daily consumption of drinking water for an adult.

The calculated MAC for antimony in drinking water is therefore 0.004 mg/L, which is below the practical quantitation level. The maximum acceptable concentration (MAC) for antimony in drinking water has therefore been set at the practical quantitation level of 0.006 mg/L.

Because antimony may leach from materials in the distribution system, sampling should be carried out on flushed samples at the point of consumption.

References