BARIUM AND BARIUM COMPOUNDS

Inductively coupled plasma–atomic emission spectrometry is a relatively effective and sensitive method for measuring low levels of barium in water, blood, urine, and bones. Detection limits of 0.25 mg barium/litre of urine, 0.6 mg barium/litre of blood, and 0.0005 mg barium/g of bone have been achieved. However, in a given sample containing barium, there is potential for interference from spectral bands of other compounds (e.g., boric acid or sodium borate) that may be present. Detection limits of 7 µg barium/litre of erythrocytes and 66 µg barium/litre of plasma have been obtained using neutron activation analysis (ATSDR, 1992).

Barium is the 16th most abundant non-gaseous element of the Earth’s crust, constituting approximately 0.04% of it. The two most prevalent naturally occurring barium ores are barite (barium sulfate) and witherite (barium carbonate). Barite occurs largely in sedimentary formations, as residual nodules resulting from weathering of barite-containing sediments, and in beds along with fluorspar, metallic sulfides, and other minerals. Witherite is found in veins and is often associated with lead sulfide. Barium is found in coal at concentrations up to 3000 mg/kg, as well as in fuel oils (IPCS, 1990; ATSDR, 1992). Estimates of terrestrial and marine concentrations of barium are 250 and 0.006 g/tonne, respectively (Considine, 1976).

Barite ore is the raw material from which nearly all other barium compounds are derived. Barite is mined in Morocco, China, India, and the United Kingdom. Crude barite ore is washed free of clay and other impurities, dried, and then ground before use. Barite is usually imported as crude ore or crushed ore for milling or as ready-milled ore. Barite can be 90–98% barium sulfate. World production of barite in 1985 was esti mated to be 5.7 million tonnes.

Because of its high specific gravity, low abrasiveness, chemical stability, and lack of magnetic effects, barite is used as a weighting agent for oil and gas well drilling muds, which counteracts high pressures encountered in the substrata (IPCS, 1990). It is also used as a filler in a range of industrial coatings, as a dense filler in some plastics and rubber products, in brake linings, and in some sealants and adhesives. The use dictates the source of barite used. Some sources produce very pure white barite, which is used in coatings, while barite from other sources is off-white and is used in applications where the colour is unimportant. The use will also dictate the particle size to which barite is milled. For example, drilling muds are ground to an average particle diameter of 44 µm, with a maximum of 30% of particles less than 6 µm in diameter. Barium and its compounds are used in diverse industrial products ranging from ceramics to lubricants. Barium is used in the manufacture of alloys, soap, rubber, and linoleum; in the manufacture of valves; as a loader for paper; and as an extinguisher for radium, uranium, and plutonium fires. Barium compounds are used in cement, specialty arc welding, glass industries, electronics, roentgenography, cosmetics, pharmaceuticals, inks, and paints. They have also been used as insecticides and rodenticides (e.g., barium metaborate, barium polysulfide, and barium fluorosilicate).

Anthropogenic sources of barium are primarily industrial. Emissions may result from mining, refining, or processing of barium minerals and manufacture of barium products. Barium is released to the atmosphere during the burning of coal, fossil fuels, and waste. Barium is also discharged in wastewater from metallurgical and industrial processes. Deposition on soil may result from human activities, including the disposal of fly ash and primary and secondary sludge in landfills (IPCS, 1990). Estimated releases of barium and barium compounds to the air, water, and soil from manufacturing and processing facilities in the USA during 1998 were 900, 45, and 9300 tonnes, respectively.²

Both specific and non-specific adsorption of barium onto oxides and soils have been observed. Specific sorption occurs onto metal oxides and hydroxides. Adsorption onto metal oxides probably acts as a control over the concentration of barium in natural waters. Electrostatic forces account for a large fraction of the non-specific sorption of barium on soil and subsoil. The retention of barium, like that of other alkaline earth cations, is largely controlled by the cation-exchange capacity of the sorbent. Complexation by soil organic material occurs to a limited extent. The K_d (soil sorption) value, the dissociation constant between sediment and barium in sediments, is 5.3 × 10⁵ ml/g (McComish & Ong, 1988).

Examination of dust falls and suspended particulates indicates that most contain barium. The presence of barium is mainly attributable to industrial emissions, especially the combustion of coal and diesel oil and waste incineration, and may also result from dusts blown from soils and mining processes. Barium sulfate and carbonate are the forms of barium most likely to occur in particulate matter in the air, although the presence of other insoluble compounds cannot be excluded. The residence time of barium in the atmosphere may be several days, depending on the particle size. Most of these particles, however, are much larger than 10 µm in size and rapidly settle back to earth. Particles can be removed from the atmosphere by rainout or washout wet deposition.

Soluble barium and suspended particulates can be transported great distances in rivers, depending on the rates of flow and sedimentation. Cartwright et al. (1978) studied the chemical control of barium solubility and showed that, for most water samples, barium ion concentration is controlled by the amount of sulfate ion in the water.

While some barium in water is removed by precipitation, exchange with soil, or other processes, most barium in surface waters ultimately reaches the ocean. Once freshwater sources discharge into seawater, barium and the sulfate ions present in salt water form barium sulfate. Due to the relatively higher concentration of sulfate present in the oceans, only an estimated 0.006% of the total barium brought by freshwater sources remains in solution (Chow et al., 1978). This estimate is supported by evidence that outer-shelf sediments have lower barium concentrations than those closer to the mainland.

Marine concentrations of barium generally increase with depth, suggesting that barium may be incorporated into organisms in the euphotic zone and subsequently sedimented and released in deeper waters (IPCS, 1990). In laboratory testing, the uptake of barium by algae in culture media was 30–60% after 15 days of exposure to barium concentrations of 0.04, 0.46, and 4.0 mg/litre of medium, the relative accumulation being inversely related to the barium concentration in the medium and directly related to the exposure duration (Havlik et al., 1980). Barium was not incorporated into organic components but was bound primarily to the cell membrane or other non-extractable components. Accumulation of barium ions (¹³³Ba) in the cells of the alga Scenedesmus obliquus has been shown to increase with increasing pH between pH 4 and 7, then remain constant over the pH range 7–9 at a barium concentration of 10^–6 mol/litre, with a calculated affinity constant (K_m) of 4.8 (Stary et al., 1984). In a marine environment contaminated with heavy metals (including barium), Guthrie et al. (1979) measured barium concentrations of 7.7 mg/litre in water and 131.0 mg/kg wet weight in sediment. Among barnacles, crabs, oysters, clams, and polychaete worms tested for barium content in this marine environment, only barnacles showed higher concentrations of barium (40.5 mg/kg wet weight) than that of the water.

Barium sulfate is present in soil through the natural process of soil formation; barium concentrations are high in soils formed from limestone, feldspar, and biotite micas of the schists and shales (Clark & Washing ton, 1924). When soluble barium-containing minerals weather and come into contact with solutions containing sulfates, barium sulfate is deposited in available geological faults. If there is insufficient sulfate to combine with barium, the soil material formed is partially saturated with barium. In soil, barium replaces other sorbed alkaline earth metals from manganese dioxide, silicon dioxide, and titanium dioxide under typical environmental conditions, by ion exchange (Bradfield, 1932; McComish & Ong, 1988). However, other alkaline earth metals displace barium from aluminium oxide (McComish & Ong, 1988).

Barium sulfate in soils is not expected to be very mobile because of the formation of water-insoluble salts and its inability to form soluble complexes with humic and fulvic materials. Under acid conditions, however, some of the water-insoluble barium compounds (e.g., barium sulfate) may become soluble and move into groundwater (US EPA, 1984).

Despite relatively high concentrations in soils, only a limited amount of barium accumulates in plants. Barium is actively taken up by legumes, grain stalks, forage plants, red ash (Fraxinus pennsylvanica) leaves, and black walnut (Juglans nigra), hickory (Carya sp.), and brazil nut (Bertholletia excelsa) trees; Douglas-fir (Pseudotsuga menziesii) trees and plants of the genus Astragallu also accumulate barium (IPCS, 1990). Barium has also been shown to accumulate in mushrooms (Aruguete et al., 1998). No studies of barium particle uptake from the air have been reported, although vegetation is capable of removing significant amounts of contaminants from the atmosphere. Plant leaves act only as deposition sites for particulate matter. Although levels of barium in wildlife have not been documented, barium has been found in dairy products and eggs (Gormican, 1970; IPCS, 1990), indicating that barium uptake occurs in animals.

A bioconcentration factor (BCF) for soil to plants was estimated as 0.4 (0.02 standard error of the mean [SEM]), based on samples of a variety of plant species (mean barium concentration of 29.8 mg/kg [13.7 SEM]) that were taken from a site in which the mean concentration of barium in the soil was 104.2 mg/kg (9.5 SEM) (Hope et al., 1996). Based on the ratio of barium concentration in the soil to whole-body barium concentra tion, the same authors computed bioaccumulation factors of 0.2 (0.002 SEM) for terrestrial insects, 0.02 (0.0004 SEM) for white-footed mice (Peromyscus leucopus), and 0.02 (0.0005 SEM) for hispid cotton rats (Sigmodon hispidus). Based on dissolved barium concentrations in surface water of 0.07 mg/litre (0.02 SEM) and whole-body barium concentrations of 2.1 mg/kg (0.5 SEM) in fish, measured at the same study site, a BCF of 129.0 litres/kg (13.5 SEM) was estimated. The authors also estimated mean depuration rates in white-footed mice and hispid cotton rats to be 0.4/day (0.01 SEM) and 0.2/day (0.01 SEM), respectively, indicating that barium is "lost from these receptors at a fairly rapid rate." Field data were collected during a single summer sampling event, and the authors advise caution in extrapolating the results to terrestrial systems in general.

There is no evidence that barium undergoes environmental biotransformation other than as a divalent cation (IPCS, 1990).

The levels of barium in air are not well docu mented, and in some cases the results are contradictory. Tabor & Warren (1958) detected barium concentrations ranging from <0.005 to 1.5 mg/m³ in the air in 18 cities and 4 suburban areas in the USA. No distinct pattern between ambient levels of barium in the air and the extent of industrialization was observed. In general, however, higher concentrations were observed in areas where metal smelting occurred (Tabor & Warren, 1958; Schroeder, 1970). In a more recent survey in the USA, ambient barium concentrations ranged from 0.0015 to 0.95 mg/m³ (US EPA, 1984). In three communities in New York City, USA, barium was measured in dust fall and household dust (Creason et al., 1975). With standard methods (US EPA, 1974), the dust fall was found to contain an average of 137 mg barium/g, while the house dust contained 20 mg barium/g.

Barium is found in almost all surface waters that have been examined (NAS, 1977). The concentrations are extremely variable and depend on local geology, water treatment, and water hardness (NAS, 1977). Barium concentrations of 7–15 mg/litre and 6 mg/litre have been measured in fresh water and seawater, respectively (Schroeder et al., 1972). The mean barium content of various US surface waters ranges from 43 to 57 mg/litre (Durum, 1960; Kopp, 1969; Kopp & Kroner, 1970; Schroeder, 1970; Bradford, 1971). The concentrations of barium in sediments of the Iowa River, USA, were measured to be 450–3000 mg/kg (Tsai et al., 1978), suggesting that barium in the water is removed by precipitation and silting.

Studies of drinking-water quality in cities in the USA have revealed levels of barium ranging from trace to 10 mg/litre (Durfor & Becker, 1964; Barnett et al., 1969; McCabe et al., 1970; McCabe, 1974; Calabrese, 1977; AWWA, 1985). Drinking-water levels of at least 1000 mg barium/litre have been reported when the barium is present mainly in the form of insoluble salts (Kojola et al., 1978). Levels of barium in Canadian water supplies have been reported to range from 5 to 600 mg/litre (Subramanian & Meranger, 1984), and barium concentrations ranging from 1 to 20 mg/litre have been measured in municipal water in Sweden (Reeves, 1986).

The concentration of barium in seawater varies greatly among different oceans and varies with factors such as latitude and depth within a given ocean. Several studies have shown that the barium content in the open ocean increases with the depth of water (Chow & Gold berg, 1960; Bolter et al., 1964; Turekian, 1965; Chow & Patterson, 1966; Anderson & Hume, 1968). A Geosecs III study of the south-west Pacific by Bacon & Edmond (1972) found a barium profile of 4.9 mg/litre in surface waters to 19.5 mg/litre in deep waters. Later studies by Chow (1976) and Chow et al. (1978) corroborated these values. Measured barium concentrations in the north-east Pacific ranged from 8.5 to 32 mg/litre (Wolgemuth & Broecker, 1970). Bernat et al. (1972) found that barium concentration profiles for the eastern Pacific Ocean and the Mediterranean Sea ranged from 5.2 to 25.2 mg/litre and from 10.6 to 12.7 mg/litre, respec tively. Anderson & Hume (1968) reported concentrations in the Atlantic Ocean ranging from 0.8 to 37.0 mg/litre in the equatorial region and from 0.04 to 22.8 mg/litre in the North Atlantic, with mean values of 6.5 and 7.6 mg/litre, respectively. In Atlantic Ocean waters off Bermuda, barium concentrations of 15.9– 19.1 mg/litre have been measured (Chow & Patterson, 1966).

The background level of barium in soils is considered to range from 100 to 3000 mg/kg, with an average of 500 mg/kg (Brooks, 1978).

Various studies document concentrations of barium in Brazil nuts ranging from 1500 to 3000 mg/kg (Robinson et al., 1950; Smith, 1971a). Barium is also present in wheat, although most is concentrated in the stalks and leaves rather than in the grain (Smith, 1971b). Tomatoes and soybeans also concentrate soil barium; the BCF ranges from 2 to 20 (Robinson et al., 1950). Levels of barium found in other food items range from <0.2 mg/kg in meats to 27 mg/kg in dry tea bags (Gormican, 1970). McHargue (1913) reported that the barium content of dry tobacco leaves was in the range of 88–293 mg/kg. Later measurements yielded 24–170 mg/ kg, with an average value of 105 mg/kg (Voss & Nicol, 1960). Most of this barium is likely to remain in the ash during burning. The concentrations of barium in tobacco smoke have not been reported.

The most important route of exposure to barium appears to be ingestion of barium through drinking-water and food. Particles containing barium may be inhaled into the lung, but little is known regarding the absorption of barium by this route.

Schroeder et al. (1972) estimated that the mean daily intake of barium is 1.24 mg in food. Hamilton & Minski (1972) estimated the total intake of barium from the diet to be 603 µg/day. The ICRP (1974) estimated barium intake from dietary sources to be approximately 0.67 mg/day. WHO (1996) reported daily dietary intake of barium for adults for the period 1970–1991 as 0.18 (minimum), 0.30 (median), and 0.72 (maximum) mg/ person. In a number of dietary studies, the average intake of barium ranged from 0.3 to 1.77 mg/day (Tipton et al., 1966, 1969; Gormican, 1970; ICRP, 1974). This is equivalent to 0.004–0.025 mg barium/kg body weight per day, assuming a 70-kg adult body weight. The barium content in school lunches from 300 schools in 19 states in the USA ranged from 0.09 to 0.43 mg/ lunch, with a mean of 0.17 mg/lunch (Murphy et al., 1971).

The barium content in drinking-water seems to depend on regional geochemical conditions. In a study of the water supplies of the 100 largest cities in the USA, a median value of 0.43 mg/litre was reported; 94% of all determinations were <0.100 mg/litre (Durfor & Becker, 1964). Assuming daily water consumption of 2 litres/person, this represents an average intake of <0.200 mg barium/day. More recent studies by Letkiewicz et al. (1984) indicated that approximately 214 million people in the USA using public water supplies are exposed to barium levels ranging from 0.001 to 0.020 mg/litre. In certain regions of the USA, however, barium levels may reach 10 mg/litre, and the average intake could be as high as 20 mg/day (Calabrese, 1977). Levels of barium in municipal water in Sweden as high as 20 mg/litre have been reported (Reeves, 1986).

Due to the paucity of information on the levels of barium in ambient air, it is difficult to estimate the intake from this source. The levels of barium in air rarely exceed 0.05 mg/m³ (Tabor & Warren, 1958). This value can be used to estimate daily barium intake via the lungs. Assuming that the average lung ventilation rates for newborn babies, male adults undergoing light activity, and male adults undergoing heavy activity are 0.5, 20, and 43 litres/min, respectively (ICRP, 1974; IPCS, 1994), the intake via inhalation would range from 0.04 to 3.1 mg/day. Other age groups and females are included in this range. Earlier, the ICRP (1974) reported that intake of barium through inhalation ranges from 0.09 to 26 mg/day. Using 0.95 mg/m³ (the upper-end estimate of ambient barium concentrations from US EPA, 1984) and the ventilation rates of ICRP (1974) for babies and adult males, a range of intakes via inhalation of 0.68–59 mg/day can be estimated.

The ICRP (1974) reported the total dietary intake of barium to be 0.75 mg/day, including both food and fluids. Schroeder et al. (1972) estimated a total of 1.33 mg/day, including food, water, and air (0.001 mg) intake.

Available data from industry in the United King dom indicate that airborne exposure to barium sulfate can range from 3.5 to 9.1 mg/m³ (8-h time-weighted average [TWA], total inhalable dust) during the manual addition of barite to mixing hoppers in the oil drilling industry, with short-term (10-min TWA, total inhalable dust) exposures as high as 34.1 mg/m³. During the processing of barite ore, in industries that typically use enclosed processes and local exhaust ventilation (LEV), exposures usually ranged between 1.3 and 3.7 mg/m³ (total inhalable dust), with highest values in one factory reaching 55.4 mg/m³. Exposure levels in the formulation of plastics and coatings, where the process is usually enclosed and LEV is used, are in the region of 1– 3.5 mg/m³ (Ball et al., 1997).

The wiring used in some speciality arc welding processes has been shown to contain 20–40% soluble barium compounds, and fumes produced during these processes contain 25% barium (Dare et al., 1984). Welders using such wire are exposed to estimated airborne concentrations of 2.2–6.2 mg soluble barium/m³ (NIOSH, 1978).

Personal air sampling in the vicinity of oven-charger and batch-mixer workers in art glass manufacturing plants revealed median ambient air concentrations of 0.041 and 0.0365 mg barium/m³, respectively (Apostoli et al., 1998). Mean concentrations of barium measured by personal sampling methods in various locations within ceramic factories in Spain ranged from 0.0012 to 0.0758 mg/m³ (Roig-Navarro et al., 1997).

Data from industry in the United Kingdom and predictions made using the Estimation and Assessment of Substance Exposure (EASE)³ model suggest that exposures to barium sulfate can be controlled to less than 10 mg/m³ 8-h TWA (total inhalable dust). In some situations, control will be to levels significantly below this value. Short-term exposures may be higher than this for some tasks.

EASE Version 2 predicts that during manual addition of barite to mixing hoppers, exposure to barium would be 2–5 mg/m³ with LEV and 5–50 mg/m³ without LEV; during dry crushing and grinding, 2–10 mg/m³ with LEV and 50–200 mg/m³ without LEV; and during dry manipulation in plastics formulation, in the range 2–5 mg/m³ with LEV and 5–50 mg/m³ without LEV. These predictions are consistent with the data from industry.

Barium sulfate is the major barium compound used in medicinal diagnostics; it is employed as an opaque contrast medium for roentgenographic studies of the gastrointestinal tract, providing another possible source of human exposure to barium (IPCS, 1990).

Information on the gastrointestinal absorption of barium in humans is limited. Lisk et al. (1988) reported the results of a mass balance study of one man who consumed a single dose of 179.2 mg barium (species not reported) in 92 g of Brazil nuts; it was estimated that at least 91% of the dose was absorbed. Barium excreted in the urine was 1.8 and 5.7% of the total dietary barium in two subjects studied by Tipton et al. (1969). Thirty-seven people were each administered a single dose (between 88 and 195 µg of barium) of one of five barium sulfate X-ray contrast media (Clavel et al., 1987). In 24 h, the total amount of barium collected in the urine ranged from 18 to 35 µg and showed a positive correlation with the amount of barium ingested. The eliminated barium was stated to be in the range 0.16– 0.26 µg/g of barium administered. Another study also indicated that a very small proportion of barium sulfate was absorbed after ingestion of barium sulfate as a radiopaque (Mauras et al., 1983).

A wide range of absorption efficiencies has been reported in animal studies. The range of reported oral absorption for all animal studies was 0.7–85.0%. This large variation may be explained in part by differences in study duration (length of time that gastrointestinal absorption was monitored), species, age, and fasting status of the animals; however, these experimental parameters did not affect gastrointestinal absorption of barium consistently among the different studies. The presence of food in the gastrointestinal tract appears to decrease barium absorption, and barium absorption appears to be higher in young animals than in older ones (US EPA, 1998).

Richmond et al. (1960, 1962a,b) studied the gastrointestinal absorption of barium chloride in several animal species. Gastrointestinal absorption was approximately 50% (barium chloride) in beagle dogs compared with 30% (barium sulfate) in rats and mice. Using the 30-day retention data from a study by Della Rosa et al. (1967), Cuddihy & Griffith (1972) estimated gastro intestinal absorption efficiencies of 0.7–1.5% in adult beagle dogs and 7% in younger beagle dogs (43– 250 days of age).

McCauley & Washington (1983) and Stoewsand et al. (1988) compared absorption efficiencies of several barium compounds. Barium sulfate and barium chloride were absorbed at "nearly equivalent rates" (based on blood and tissue levels) in rats following a single gavage dose of similar barium concentrations (McCauley & Washington, 1983). Similar concentrations of barium were found in the bones of rats fed diets with equivalent doses of barium chloride or barium from Brazil nuts. McCauley & Washington (1983) suggested that the similarity in absorption efficiency between barium sulfate and barium chloride may have been due to the ability of hydrochloric acid in the stomach to solubilize small quantities of barium sulfate (barium chloride, barium sulfate, or barium carbonate had been adminis tered to the rats at a concentration of 10 mg ¹³³Ba/litre in the drinking-water at pH 7.0). This is supported by the finding that barium carbonate in a vehicle containing sodium bicarbonate was poorly absorbed. The buffering capacity of sodium bicarbonate may have impaired the hydrochloric acid-mediated conversion of barium carbonate to barium chloride. The results of these studies suggest that soluble barium compounds and/or barium compounds that yield a dissociated barium ion in the acid environment of the upper gastrointestinal tract have similar absorption efficiencies.

There is no direct evidence in humans that barium is absorbed by the respiratory tract. However, Zschiesche et al. (1992) reported increased plasma and urine levels of barium compounds in workers exposed to barium during welding, thus indicating that airborne barium is absorbed either by the respiratory system or by the gastrointestinal tract following mucociliary clearance. Following termination of barite exposure, Doig (1976) showed a clearing of lung opacities in workers.

A suspension of 23, 233, or 2330 mg ¹³³Ba in isotonic saline was instilled into the trachea and then blown into the "deep respiratory tract" of rats (Cember et al., 1961). Four rats from each group were sacrificed at intervals up to 20 days after administration, and lungs, kidneys, spleen, and tracheobronchial lymph nodes were extracted and examined radiologically for the presence of the ¹³³Ba. The clearance half-time of the ¹³³Ba from the deep respiratory tract for all dose levels was determined to be between 8 and 10 days and was not influenced by the dose administered. Less than 0.1% of the instilled dose of ¹³³Ba was detected in the tissues analysed (excluding lungs).

Animal studies provide evidence that barium compounds, including poorly water soluble barium sulfate, are cleared from the respiratory tract. Collectively, these studies suggest that barium is absorbed following inhalation exposure. Morrow et al. (1964) estimated that the biological half-time of ¹³¹BaSO₄ in the lower respiratory tract was 8 days in dogs inhaling 1.1 mg barium sulfate/ litre for 30–90 min. Twenty-four hours after an intratracheal injection of ¹³³BaSO₄, 15.3% of the radioactivity was cleared from the lungs. The barium sulfate was cleared via mucociliary clearance mechanisms (7.9% of initial radioactive burden) and via lung-to-blood transfer (7.4% of radioactivity) (Spritzer & Watson, 1964). Clearance half-times of 66 and 88 days were calculated for the cranial and caudal regions of the trachea in rats intratracheally administered 2 mg ¹³³BaSO₄ (Takahashi & Patrick, 1987). Cuddihy et al. (1974) showed uptake of barium in the bone following inhalation exposure in rats.

Differences in water solubility appear to account for observed differences in respiratory tract clearance rates for barium compounds. The clearance half-times were proportional to solubility in dogs exposed to aerosols of barium chloride, barium sulfate, heat-treated barium sulfate (likely oxidized), or barium incorporated in fused montmorillonite clay particles (Cuddihy et al., 1974).

No data are available on dermal absorption of barium compounds.

The highest concentrations of barium (approximately 91% of the total body burden) are found in the bone (IPCS, 1990). Reeves (1986) noted that osseous uptake of barium was 1.5–5 times higher than that of calcium or strontium. In the bone, barium is primarily deposited in areas of active bone growth (IPCS, 1990). The uptake of barium into the bone appears to be rapid. One day after rats were exposed to barium chloride aerosols, 78% of the total barium body burden was found in the skeleton; by 11 days post-exposure, more than 95% of the total body burden was found in the skeleton (Cuddihy et al., 1974).

The remainder of the barium in the body is found in soft tissues, particularly aorta, brain, heart, kidney, spleen, pancreas, and lung (IPCS, 1990). High concentrations of barium are sometimes found in the eye, primarily in the pigmented structures (Reeves, 1986). McCauley & Washington (1983) found that 24 h after administration of an oral dose of ¹³³BaCl₂ to dogs, ¹³³Ba levels in the heart were 3 times higher than in the eye, skeletal muscle, and kidneys, which had similar concentrations. Levels in these tissues were higher than the whole-blood concentration, suggesting that they con centrated barium.

Barium is excreted primarily in the faeces following oral, inhalation, and parenteral exposure, but it is also excreted in the urine. At a normal intake level of 1.33 mg barium/day (1.24, 0.086, and 0.001 mg/day from food, water, and air, respectively), humans eliminated approximately 90% of the barium in the faeces and 2% in the urine (Schroeder et al., 1972). Tipton et al. (1969) found similar results; in two men studied, 95–98% and 2–5% of the daily barium intake were excreted in the faeces and urine, respectively. In the tracheal instillation study of Cember et al. (1961), urine and faeces were collected for 21 days in two high-dose animals. Faecal elimination accounted for around two-thirds of the radioactivity administered, and the urine for around 10%. Overall, this study indicated that very little of the administered barium is absorbed, with the majority of the compound being eliminated in the faeces.

The biological half-times of barium of 3.6, 34.2, and 1033 days were estimated in humans using a three-component exponential function (Rundo, 1967). Following inhalation exposure to ¹⁴⁰BaCl₂–¹⁴⁰LaCl₂, a half-time of 12.8 days was estimated in beagle dogs (Cuddihy & Griffith, 1972).

Acute oral LD₅₀ values in rats for barium chloride, barium carbonate, and barium sulfide range from 118 to 800 mg/kg body weight (IPCS, 1990; ATSDR, 1992). Acute effects include fluid accumulation in the trachea, intestinal inflammation, decreased liver/brain weight ratio, darkened liver, increased kidney/body weight ratio, and decreased body weight (Borzelleca et al., 1988). No data were available regarding the lethality of barium sulfate in laboratory animals.

In rabbits administered single intratracheal doses of ¹⁴⁷Ba (85% barium sulfate) in the range of 0.015– 0.6 ml/kg body weight, soft X-rays of the lungs revealed dose-related shadows, and transient bronchopneumonia, bronchitis, or bronchiolitis was observed (Uchiyama et al., 1995).

Barium sulfate instilled in the trachea or bronchus of laboratory animals was present in the lungs for up to 126 days after administration and induced local increases in the number of polymorphonuclear leukocytes 1 day post-instillation, followed by an increase in macrophages. Small foci of atelectasis or emphysema were seen post-instillation, whereas hyperplasia and/or granulomas in bronchial tissue (without evidence of pulmonary fibrosis) appeared between days 7 and 42 post-instillation (Huston & Cunningham, 1952; Willson et al., 1959; Nelson et al., 1964; Stirling & Patrick, 1980; Ginai et al., 1984; Slocombe et al., 1989). None of the investigators reported systemic effects.

Intravenous infusion of barium chloride into anaesthetized dogs (0.5–2 µmol/kg body weight per minute) or guinea-pigs (1.7 mg/kg body weight per minute) resulted in increased blood pressure and cardiac arrhythmias (Roza & Berman, 1971; Hicks et al., 1986). The study in dogs also reported skeletal muscle flaccidity and paralysis (Roza & Berman, 1971). Determination of plasma potassium concentrations in the dogs revealed severe hypopotassaemia, which was attributed to an extracellular-to-intracellular shift of potassium. Simultaneous infusion of potassium into the dogs abolished the cardiac effects and the skeletal muscle flaccidity but did not affect hypertension. The hypertension did not appear to be mediated through the renin–angiotensin system, because it was not prevented by bilateral nephrectomy of the dogs. Dose-dependent cardiac arrhythmias were also noted in conscious rabbits following infusion of barium chloride (Mattila et al., 1986).

Barium hydroxide is strongly alkaline and therefore corrosive. Topical and ocular applications (24-h exposure) of barium nitrate and barium oxide in rabbits caused mild skin irritation and severe eye irritation (RTECS, 1985). No data are available on skin or eye irritation caused by barium sulfate. However, the physicochemical properties of barium sulfate and the lack of reports of skin or eye irritation in humans despite its widespread use, particularly for X-ray purposes, suggest that barium sulfate is not irritating or corrosive to either skin or eyes.

Useful information on the sensitization potential of barium compounds was not identified.

Increased blood pressure was reported in rats exposed to barium chloride in drinking-water for 1 month at an estimated daily dose of 7.1 mg barium/kg body weight (Perry et al., 1983, 1985, 1989). No chemically related adverse effects were seen in rats exposed to up to 2000 mg barium chloride dihydrate/litre in drinking-water (average daily doses of up to 110 mg barium/kg body weight) for 15 days. In mice similarly exposed to up to 692 mg/litre (average daily doses of up to 70 and 85 mg barium/kg body weight in males and females, respectively), the only significant adverse effect was an increased relative liver weight in high-dose males (NTP, 1994).

Muller (1973) exposed rats to barium sulfate dust at an exposure level of 40 mg/m³ (particle size 1–2 µm), 5 h/day, 5 days/week, for up to 8 weeks. Following a single exposure period, thickening of the alveolar septa, loss of ciliated epithelial cells, and formation of multi cellular epithelium were noted. At 14 days of treatment, rats exhibited normal alveolar septa. However, the investigator reported unspecified changes in bronchiolar epithelium that were still present following a 28-day recovery period.

NTP (1994) treated groups of rats (10 per sex per group) with barium chloride dihydrate in drinking-water at concentrations of 0, 125, 500, 1000, 2000, or 4000 mg/litre for 13 weeks (average daily doses of 0, 10, 30–35, 65, 110–115, or 180–200 mg barium/kg body weight). Effects observed in the 4000 mg/litre rats included reduced water consumption, significantly reduced final mean body weights, and death of three males and one female during the last week of the study. It may be noted that the 4000 mg/litre (180–200 mg barium/kg body weight) dose is comparable to the LD₅₀ of 180 mg barium/kg body weight in rats. There were no clearly chemical-related clinical findings of toxicity or cardiovascular (heart rate, systolic blood pressure, electrocardiogram) effects. Toxicologically significant (P æ 0.01) organ weight changes consisted of increased absolute and relative kidney weights in 2000 and 4000 mg/litre female rats, increased relative kidney weights in 4000 mg/litre male rats, and decreased absolute and/or relative liver weights in 4000 mg/litre rats of both sexes. Organ weight changes in the kidney were considered to be associated with chemical-induced renal lesions consisting of minimal to mild, focal to multifocal areas of dilatation of the proximal convoluted tubules seen in three rats of each sex at the 4000 mg/litre exposure level. Crystals were not present in the kidney tubules. Decreased liver weights and lymphoid depletions in spleen, thymus, and/or lymph nodes of 4000 mg/litre rats were attributed to reduced body weight and stress. There were no biologically significant changes in serum electrolytes or haematology values that were considered to be chemical related. Significant decreases in the magnitude of undifferentiated motor activity were observed at day 90 in 4000 mg/litre rats of both sexes. Marginal decreases in undifferentiated motor activity were seen in all other barium-exposed groups except the 1000 mg/litre female rats. No significant or dose-related changes were observed in other neuro behavioural end-points. Although NTP (1994) con sidered the no-observed-adverse-effect level (NOAEL) to be 115 mg/kg body weight per day (the 2000 mg/litre exposure level), US EPA (1998) suggested that this level might be considered a lowest-observed-adverse-effect level (LOAEL), based on significant (P æ 0.01) increased kidney weight in female rats and an observed LD₅₀ of 118 mg/kg body weight in rats (RTECS, 1985). The NOAEL would then be 65 mg/kg body weight per day (the 1000 mg/litre exposure level).

NTP (1994) also treated groups of mice (10 per sex per group) with barium chloride dihydrate in drinking-water at concentrations of 0, 125, 500, 1000, 2000, or 4000 mg/litre for 13 weeks (average daily doses of 0, 15, 55–60, 100–110, 200–205, or 450– 495 mg barium/kg body weight). Adverse effects in the 4000 mg/litre groups included death of 6 males and 7 females, chemical-related nephropathy in 10 males and 9 females, significantly reduced body weights in both sexes, reduced absolute kidney weight in males, and increased relative kidney weight in females, relative to controls. Kidney lesions were characterized by tubule dilatation, renal tubule atrophy, tubule cell regeneration, and the presence of crystals, primarily in the lumen of the renal tubules. Relative and absolute thymus weights were decreased in both sexes. Lymphoid depletions in spleen, thymus, and/or lymph nodes of 4000 mg/litre mice were attributed to reduced body weight and stress. A significant decrease in forelimb grip strength of 4000 mg/litre female mice, observed at 90 days, was attributed to debilitation; no significant dose-related changes were observed in other neurobehavioural end-points. Cardiovascular tests were not performed in mice. The LOAEL is 495 mg/kg body weight per day, based on nephropathy and mortality at the 4000 mg/litre exposure level; the NOAEL is 205 mg/kg body weight per day.

Tardiff et al. (1980) exposed male and female Charles River rats continuously to barium chloride in drinking-water for up to 13 weeks. The authors estimated doses as 0, 1.7, 8.1, or 38.1 mg barium/kg body weight per day for males and 0, 2.1, 9.7, or 45.7 mg barium/kg body weight per day for females. Rats were fed a diet of Tekland mouse/rat diet pellets, which contributed a baseline dose of 0.5 µg barium/kg body weight per day. The only reported adverse effects were depressed water consumption in the high-dose groups of both sexes and slight decreases in relative adrenal weights in mid-dose males at 8 weeks and in all exposed groups of females at 13 weeks; these changes were not dose related. Blood pressure and end-points sensitive for glomerular damage (electron microscopic examination or urinary excretion of protein) were not investigated.

In a series of longer-term histological, electron microscopic, electrocardiographic, and blood pressure studies (McCauley et al., 1985), CD Sprague-Dawley rats were given barium in drinking-water for various durations and fed Purina rat chow (containing 12 mg barium/kg) or Tekland rat chow (insignificant barium intake). In the histology studies, three exposure regimens were used with the Purina rat chow diet, and the estimated total barium intakes were 1, 1.15, 2.5, 16, or 38.5 mg/kg body weight per day for 36–68 weeks. Histological evaluations of an extensive number of tissues did not reveal barium-related lesions. No alterations in haematocrit levels were observed. A retinal lesion ("focal absence of the outer layers of the retina") was observed but did not appear to be dose or duration related; its relationship to barium exposure is uncertain. No significant increases in incidences of neoplasms were observed in the barium-exposed rats, but the study duration is less than a lifetime and may not have been of sufficient duration for the detection of late-developing tumours.

In a 16-week blood pressure study (McCauley et al., 1985), normotensive rats were fed Tekland rat chow (0.5 mg barium/kg body weight per day) and administered barium in drinking-water or in 0.9% sodium chloride solution as drinking-water (estimated daily barium doses of 0, 0.45, 1.5, 4.5, or 15 mg/kg body weight). Unilaterally nephrectomized rats were similarly treated (estimated daily doses of 0.15, 1.5, 15, or 150 mg barium/kg body weight), while groups of Dahl salt-sensitive and Dahl salt-resistant rats received esti mated daily doses of 0.15, 1.5, 15, or 150 mg barium/kg body weight in 0.9% sodium chloride as drinking-water. The authors stated that all groups showed fluctuations of blood pressure. No indications of hypertension were observed, but there were no 0 mg barium/litre / 0.9% sodium chloride controls in the study. Electron microscopic examination of kidneys in all the rats in the blood pressure studies demonstrated no changes in arteriolar vessel walls or in tubules of the nephrons. However, structural changes in glomeruli were observed in the high-dose (150 mg barium/kg body weight per day) nephrectomized, Dahl salt-sensitive, and Dahl salt-resistant groups. No glomerular effects were seen at the next lower exposure level in any group of rats.

Data on the toxicity of barium compounds in animals following inhalation exposure are limited to a study in which male albino rats were exposed to barium carbonate at 0, 1.15, or 5.20 mg/m³ (0, 0.80, or 3.6 mg barium/m³) for 4 h/day, 6 days/week, for 4 months (Tarasenko et al., 1977). At 5.20 mg/m³ (but not 1.15 mg/m³), reported alterations included a 21% decrease in body weight gain, a 32% increase in arterial pressure, altered haematological parameters, altered serum chemistry parameters, increased calcium levels in the urine, impaired liver function, and histological alterations in the heart, liver, kidneys, and lungs. The authors noted that the heart, liver, and kidneys "had a character of mild protein (‘granular’) dystrophy."

NTP (1994) treated male and female F344/N rats (60 animals per dose group per sex) with deionized drinking-water containing 0, 500, 1250, or 2500 mg barium chloride dihydrate/litre for 2 years. Daily doses of barium were estimated to be 0, 15, 30, or 60 mg/kg body weight for males and 0, 15, 45, or 75 mg/kg body weight for females. The animals were fed an NIH-07 mash diet; the barium content of the diet was not reported. In this study, neurobehavioural and cardiovascular tests were not performed. Reported exposure-related effects included reduced body weights in some mid-and high-dose rats, dose-related decreased water consumption, and significantly increased relative kidney weights in high-dose females (the only indication of potential adverse renal effects). Thus, the 2500 mg/litre exposure level (60 mg barium/kg body weight per day for males and 75 mg barium/kg body weight per day for females) may be a chronic NOAEL or LOAEL for rats, depending on interpretation of the increased relative kidney weight in females. When considered together with the results in the 13-week NTP (1994) study in rats, in which increased relative and absolute kidney weights were seen in female rats receiving 2000 mg barium/litre in drinking-water (115 mg barium/kg body weight per day) and kidney lesions accompanied by increases in relative and absolute kidney weights were seen in female rats at 4000 mg/litre (180 mg barium/kg body weight per day), the increased relative kidney weight in females of the 2-year study is suggestive of potential renal effects. Therefore, 75 mg barium/kg body weight per day is designated a chronic LOAEL and 45 mg barium/kg body weight per day a chronic NOAEL for female rats for renal effects in the NTP (1994) study. There were no significant increases in incidences of neoplasms in the barium-exposed rats. Significant negative trends were observed in the incidences of mononuclear cell leukaemia in male rats, benign and malignant adrenal pheochromocytoma in male rats, and mammary gland neoplasms (fibroadenoma, adenoma, or carcinoma) in female rats.

NTP (1994) also treated B6C3F₁ mice (60 animals per dose group per sex) with drinking-water containing 0, 500, 1250, or 2500 mg barium chloride dihydrate/litre for 2 years. Estimated daily doses were 0, 30, 75, or 160 mg barium/kg body weight for males and 0, 40, 90, or 200 mg barium/kg body weight for females. The animals were fed an NIH-07 mash diet; the barium con tent of the diet was not reported. Neurobehavioural and cardiovascular tests were not performed. At the 15-month interim evaluation, the absolute and relative spleen weights of the 2500 mg/litre female mice were significantly (P < 0.01) lower than those of the controls, and the absolute and relative thymus weights of the 2500 mg/litre male mice were marginally lower than those of the controls. Additionally, survival rates for the 2500 mg/litre mice at the end of study were significantly (P < 0.01) lower than those of the controls, which was attributed to chemical-related renal lesions. These renal lesions were characterized by tubule dilatation, renal tubule atrophy, tubule cell regeneration, hyaline cast formation, multifocal interstitial fibrosis, and the presence of crystals, primarily in the lumen of the renal tubules. Lymphoid depletions in the spleen, thymus, and lymph nodes were observed in 2500 mg/litre male and female mice, particularly in animals that died early, and were thought to be the result of debilitation associated with nephropathy. Thus, the chronic LOAEL in mice is 2500 mg/litre (160 mg barium/kg body weight per day for males and 200 mg barium/kg body weight per day for females). The next lower exposure level, 1250 mg/litre (75 mg barium/kg body weight per day in males and 90 mg barium/kg body weight per day in females), is a chronic NOAEL. The incidences of neoplasms in the barium-exposed mice were not significantly higher than in control mice. In the 2500 mg/litre female mice, the incidences of several neoplasms were significantly lower than in the controls; the authors attributed this finding to the marked reduction in survival in the barium-exposed animals.

Schroeder & Mitchener (1975a) exposed Long-Evans rats (52 per sex per group) to 0 or 5 mg barium/ litre (as barium acetate) in drinking-water from weaning to natural death (approximately 2 years). Dosages from drinking-water were 0.61 mg barium/kg body weight per day for males and 0.67 mg barium/kg body weight per day for females based on reference body weights and water intakes from US EPA (1988). The diet was characterized as a "low metal" diet, and it included 60% rye flour, 30% dried skim milk, 9% corn oil, 1% iodized chloride, and assorted vitamins; the barium content was not reported. Barium had no significant effect on the growth of males, but increased the growth of older females. The incidence of proteinuria in males exposed to barium for approximately 152 days (at 173 days of age) was significantly higher than in controls. Female rats at 532 and 773 days of age had higher serum choles terol concentrations, and males at these ages had serum glucose levels different from controls; the authors attached no biological or toxicological significance to these serum chemistry results. Histopathology of heart, lung, kidney, liver, and spleen did not reveal alterations. No significant increases in the number of gross tumours were observed in the barium-exposed male or female rats.

Kopp et al. (1985) treated weanling female Long-Evans rats with barium chloride in their drinking-water (100 mg/litre) for 16 months and compared them with a control group supplied with water containing no barium. All animals received a standard rye-based diet, low in heavy metal content. Random batches of this feed were assayed for metal content and contained 1.5 µg barium/g feed. Average final body weights for both groups were found to be the same (control, 421 g; 100 mg barium/litre, 431 g). Furthermore, the measured haematological characteristics as well as feed and water consumption were not affected during the 16-month experiment. However, a significant increase in the average systemic blood pressure was detected in the barium-exposed rats after 1 month exposure and thereafter.

In similarly exposed Charles River CD white-mice (36–54 per sex) (Schroeder & Mitchener, 1975b), dosages from drinking-water were 1.18 mg barium/kg body weight per day for males and 1.20 mg barium/kg body weight per day for females (US EPA, 1988). Growth and body weights were not affected by the barium treatment. Histology of the heart, lung, liver, kidney, and spleen was normal. In males, longevity (defined as the mean life span of the last surviving five animals of each sex in each treatment group) was significantly reduced. The mean life span, however, was not affected. The incidences of lymphoma leukaemia and lung tumours in the male and female mice exposed to barium were not significantly different from the inci dences in the control mice.

Perry et al. (1983, 1985, 1989) exposed female weanling Long-Evans rats to 0, 1, 10, or 100 mg barium/ litre (as barium chloride) in drinking-water for 1, 4, and 16 months. Drinking-water was fortified with five essential metals (1 mg molybdenum/litre, 1 mg cobalt/ litre, 5 mg copper/litre, 10 mg manganese/litre, and 50 mg zinc/litre). All animals received a rye-based diet with low trace metal content based on that used by Schroeder & Mitchener (1975a,b). After 8 months of exposure to 10 mg/litre, mean systolic blood pressure had increased by 6 mmHg (800 Pa) and continued to be significantly elevated through 16 months (+4 mmHg [530 Pa]). Significant increases in mean systolic blood pressure were evident at 100 mg/litre starting at 1 month (+12 mmHg [1600 Pa]) and continuing through 16 months (+16 mmHg [2130 Pa]). An additional 12 rats, exposed for 16 months to 100 mg/litre, exhibited a reduction of ATP and phosphocreatinine content of the myocardium, depressed rates of cardiac contraction, and depressed electrical excitability (compared with 18 control rats). Since this study used a diet low in essential metals, specifically calcium, the observation of barium chloride-related effects on hypertension in rats is of questionable significance to humans.

There is a limited amount of information available on the genotoxicity of barium compounds. No in vivo studies have been conducted. Most in vitro studies have found that barium chloride and barium nitrate did not induce gene mutations in bacterial assays with or with out metabolic activation. Ames assays with Salmonella typhimurium strains TA1535, TA1537, TA1538, TA97, TA98, and TA100 with or without metabolic activation (Monaco et al., 1990, 1991; NTP, 1994), rec assays with Bacillus subtilis strains H17 and H45 (Nishioka, 1975; Kanematsu et al., 1980), and a microscreen assay with Escherichia coli with metabolic activation (Rossman et al., 1991) have produced negative results with barium chloride. Negative results have also been observed for barium nitrate in the rec assay using B. subtilis strains H17 and H45 (Kanematsu et al., 1980). Barium chloride induced gene mutations in L5178Y mouse lymphoma cells with, but not without, metabolic activation (NTP, 1994). Neither barium acetate nor barium chloride decreased the fidelity of DNA synthesis in avian myeloblastosis virus DNA polymerase (Sirover & Loeb, 1976). In mammalian cells, barium chloride did not induce sister chromatid exchanges or chromosomal aberrations in cultured Chinese hamster ovary cells, with or without activation (NTP, 1994). In summary, except for the mouse lymphoma assay, results of in vitro tests have been generally negative.

Data on the reproductive and developmental toxicity of barium compounds are limited. Decreased ovary weight and ovary/brain weight ratio were seen in female rats administered oral gavage doses of 198 mg barium/kg body weight per day, once a day for 10 days (Borzelleca et al., 1988). In single-generation reproductive toxicity studies in rats and mice (Dietz et al., 1992), groups of 20 male and 20 female F344/N rats and B6C3F₁ mice were exposed to barium chloride dihydrate in drinking-water for up to 60 days. The barium chloride dihydrate concentrations were 0, 1000, 2000, or 4000 mg/litre (estimated by the authors to be 0, 50, 100, and 200 mg/kg body weight per day) for the rats and 0, 500, 1000, or 2000 mg/litre (estimated by the authors to be 0, 50, 100, and 200 mg/kg body weight per day) for the mice. The authors measured weekly body weight changes and water consumption, which were used to estimate daily barium exposure in both mice and rats. After the completion of the exposure period (60 days), males and females from the same dosage groups were housed together until there was evidence of mating or until the end of the mating period (8 days). This rodent study reported fertility index, fetal and maternal toxicity, and developmental toxicity end-points in fetus and neonates. There were no indications of reproductive or developmental toxicity in any of the exposure groups. However, the results should be interpreted cautiously because of below-normal pregnancy rates in all groups of exposed, as well as control, rats and mice.

Ridgeway & Kanofsky (1952) examined the developmental toxicity of barium by injecting 20 mg barium chloride into the yolk sac of developing chick embryos. When injection was made on day 8 of development, developmental defects were observed in toes. In contrast, no effects were seen when injection was made on day 4 of development.

Tarasenko et al. (1977) also reported that a shortening of the mean duration of the estrous cycle and an alteration in the proportion of mature and dying ovarian follicles were observed in rats exposed to 13.4 mg bar ium carbonate/m³ (9.3 mg barium/m³) for 4 months, compared with a control group. These effects were not observed in rats exposed to 3.1 mg/m³ (2.2 mg barium/ m³). The authors also reported that rats in the 13.4 mg/ m³ group gave birth to underdeveloped offspring that showed considerable mortality and slow body weight gain during the first 2 postnatal months.

Only limited information is available on the immunotoxicity and neurotoxicity of barium compounds (IPCS, 1990). Intravenous infusion of barium chloride into anaesthetized dogs resulted in muscle flaccidity and paralysis, which appeared to result from severe hypo potassaemia (Roza & Berman, 1971).

Intentional or accidental ingestion of barium compounds (i.e., barium carbonate, barium chloride) causes gastroenteritis (vomiting, diarrhoea, abdominal pain), hypopotassaemia, hypertension, cardiac arrhythmias, and skeletal muscle paralysis. Potassium infusion is used clinically to reverse the toxic effects of barium (Diengott et al., 1964; Gould et al., 1973; IPCS, 1990; US EPA, 1990, 1998).

According to RTECS (1985), the lowest lethal acute oral doses for barium chloride and barium carbonate are 11.4 and 57 mg/kg body weight, respectively; for barium carbonate, a dose as low as 29 mg/kg body weight causes flaccid paralysis, paraesthesia, and muscle weakness.

Opacities were detected on lung X-rays of three patients for up to 2 years following accidental aspiration of barium sulfate orally administered for observation of the gastrointestinal tract (Buschman, 1991).

In a case report involving the grinding of barite ore, a worker was exposed over a period of 10 years to extremely high total dust concentrations (approximately 212 000 particles/cm³ for 1.5 h and 60 000 particles/cm³ for 1 h), although it was not stated how these measurements were made (Pendergrass & Greening, 1953). Analysis indicated that 49% of the workplace airborne dust was barium sulfate, although particle size was not stated. After 2 years of exposure to the barite ore dust, fine nodulation was observed on lung X-rays, apparently due to the presence of barium sulfate. The presence of barium sulfate in the lung tissue was confirmed by chemical analysis and light microscopy at autopsy 11 years after cessation of exposure. Several histo pathological findings were observed, including fibrosis (although mostly characteristic of silicosis). The histo pathological findings were considered to be due to the silica and anthracite exposure; the X-ray opacities seen were attributed to the presence of the barium sulfate.

In an extensive study, temperature and pulse rate measurements were taken as an indication of an acute inflammatory response for 291 humans administered a single unstated dose of a 50% w/v barium sulfate suspension for bronchographic purposes (Nelson et al., 1964). The method of administration was unstated, but the suspension was presumed to have been instilled into the trachea and then blown into the lungs. In 154 patients, there was radiological evidence of the presence of barium sulfate in the bronchial tree at the time of the last available X-ray (various time points ranging from <1 week to >1 year after administration); in 135 patients, on the other hand, there was no radiological evidence of residual barium sulfate in the lungs 1 year after bronchography. Forty-one of these patients exhibited complete elimination of the barium sulfate from the lungs within 1 week; it was stated that in some of these patients, this clearance occurred within 24 h.

Wones et al. (1990) administered 1.5 litres/day of distilled drinking-water containing various levels of barium chloride to 11 healthy male volunteers aged 27–61 years (mean 39.5 years, median 41 years). None of the subjects was taking any medications, and none had hypertension, diabetes, or cardiovascular disease. Barium concentrations in the drinking-water consumed by the subjects prior to the study were known to be very low. No barium was added for the first 2 weeks, which served as a control period; drinking-water containing 5 mg barium/litre (0.14 mg barium/kg body weight per day using reference values of 2 litres/day for water consumption and 70 kg for body weight) was administered for the next 4 weeks, and drinking-water containing 10 mg barium/litre (0.21 mg barium/kg body weight per day) was administered for the last 4 weeks of the study. Diets were controlled to mimic US dietary practices (barium content of the diet was not determined, but the authors mentioned that a typical hospital diet provides 0.75 mg barium/day, or 0.011 mg barium/kg body weight per day using a 70-kg reference weight). All beverages and food were provided, and subjects were instructed to consume only what was provided. The sub jects were also instructed to keep their level of exercise constant and to abstain from alcohol, and smokers were told to smoke consistently throughout the study. Systolic and diastolic blood pressures were measured in the morning and evening. Blood was collected at the beginning and periodically, particularly as four consecutive daily samples at the end of each of the three study periods. Twenty-four-hour urine collections were per formed at the end of each study period. Twenty-four-hour continuous electrocardiographic monitoring was performed on 2 consecutive days at the end of each study period.

Blood pressures were not significantly affected by barium exposure (Wones et al., 1990). A trend towards increased total serum calcium with barium exposure was noted but was not considered to be clinically significant. No significant changes were observed in plasma total cholesterol, triglycerides, low-density lipoprotein (LDL) or high-density lipoprotein (HDL) cholesterol, LDL:HDL ratio, and apolipoproteins A1, A2, and B. Serum glucose, albumin, and potassium levels and urinary levels of sodium, potassium, or metanephrines (catecholamine breakdown products) were unchanged. Electrocardiograms revealed no changes in cardiac cycle intervals, including the QT interval; the study authors noted that the lack of shortening of the QT interval provided evidence that the slight increase in serum calcium was not clinically significant. In addition, no significant arrhythmias, no increase in ventricular irritability, and no apparent conduction problems were seen with barium exposure. This study did not identify a LOAEL; the NOAEL is 0.21 mg barium/kg body weight per day.

Transient cell transformations resembling severe premalignant dysplasia were noted following single topical applications (four times at intervals of 4– 6 weeks) of 1.25 mmol barium chloride/litre to the cervix of a woman with no known history of abnormal cervical cytology (Ayre, 1966). In another case (Ayre & LeGuerrier, 1967), cell transformations similar to extreme dysplasia and resembling cell findings of cancer in situ were observed following a single topical appli cation of 1.25 mmol barium chloride/litre (mixed with equal amounts of 70% dimethylsulfoxide) to the cervix.

Brenniman & Levy (1984) reported an ecological epidemiological study of mortality and morbidity in populations living in communities in Illinois, USA, with elevated levels of barium in municipal drinking-water (2–10 mg/litre, 0.06–0.3 mg barium/kg body weight per day assuming water consumption of 2 litres/day and weight of 70 kg) or low levels of barium in drinking-water (0.2 mg/litre, 0.006 mg barium/kg body weight per day). Barium was the only drinking-water contaminant that exceeded drinking-water regulations of the time in any of the public drinking-water supplies. The communities were matched for demographic character istics and socioeconomic status. Communities that were industrialized or geographically different were excluded. Although the study attempted to exclude communities with high rates of population change, two of the four high-barium communities had about 75% change in population between 1960 and 1970; these were kept in the study for lack of satisfactory replacements.

In the mortality study (Brenniman & Levy, 1984), age-adjusted mortality rates for cardiovascular diseases (combined), heart diseases (arteriosclerosis), and all causes for both sexes together were significantly higher in the elevated-barium communities compared with the low-barium communities for the years 1971–1975. These differences were largely confined to the population 65 years of age or older. This study did not measure the barium exposure of individual subjects and did not control for several important variables, such as population mobility (approximately 75% turnover in two of the four high-barium communities from 1960 to 1970), use of water softeners that would remove barium from and add sodium to the water supply, use of medication by study subjects, and other risk factors, such as smoking, diet, and exercise. As a result, it is not possible to assign a causal relationship between mortality and exposure to barium.

The morbidity study (Brenniman & Levy, 1984) was conducted on two Illinois, USA, communities, McHenry and West Dundee, which had similar demographic and socioeconomic characteristics, but a 70-fold difference in barium concentrations in drinking-water. The mean concentration in McHenry’s drinking-water was 0.1 mg barium/litre, whereas the mean concentration in West Dundee’s drinking-water was 7.3 mg barium/litre. The levels of other minerals in the drinking-water of the two communities were stated to be similar. Subjects (2000) were selected randomly from a pool that included every person 18 years of age or older in a random sample of blocks within each community. All subjects underwent three blood pressure measurements (taken over a 20-min period with a calibrated electronic blood pressure apparatus) and responded to a health questionnaire that included such variables as sex, age, weight, height, smoking habits, family history, occupation, medication, and physician-diagnosed heart disease, stroke, and renal disease. Data were analysed using the signed rank test for age-specific rates, the weighted Z test for prevalence rates, and analysis of variance for blood pressures. No significant differences in mean systolic or diastolic blood pressures or in history of hypertension, heart disease, stroke, or kidney disease (which included serum and urinary protein and creatinine levels) were found for men or women of the two communities. A more controlled study was con ducted on a subpopulation of the McHenry and West Dundee subjects who did not have home water softeners, were not taking medication for hypertension, and had lived in the study community for more than 10 years. No significant differences were observed between the mean systolic or diastolic blood pressures for men or women of these subpopulations in the low-barium (0.1 mg barium/litre, 0.0029 mg barium/kg body weight per day assuming water ingestion of 2 litres/day and 70-kg body weight) and elevated-barium communities (7.3 mg barium/litre, 0.21 mg barium/kg body weight per day).

The database on the toxicity of inhaled barium compounds in humans consists primarily of studies of occupational exposure to barium sulfate or barite ore or to unspecified soluble barium compounds. Several case reports (e.g., Pendergrass & Greening, 1953; Seaton et al., 1986) and a prospective study conducted by Doig (1976) have reported baritosis in barium-exposed workers. Baritosis is considered to be a benign pneumoconiosis resulting from the inhalation of barite ore or barium sulfate. The most outstanding feature of baritosis is the intense radiopacity of the discrete opacities that are usually profusely disseminated throughout the lung fields; in some cases, the opacities may be so numerous that they appear confluent. The Third Conference of Experts on Pneumoconiosis (ACGIH, 1992) noted that barium sulfate produced a non-collagenous type of pneumoconiosis in which there is a minimal stromal reaction that consists mainly of reticulin fibres, intact alveolar architecture, and potentially reversible lesions. The available human data on baritosis suggest that the accumulation of barium in the lungs does not result in medical disability or symptomatology. A decline in the profusion and opacity density, suggesting a decrease in the amount of accumulated barium in the lung, has been observed several years after termination of exposure.

Doig (1976) reported on a series of cross-sectional examinations of workers at a barite grinding facility. During the initial investigation in 1947, five workers employed for more than 3.5 years were examined. No evidence of baritosis was observed in any of the workers. In 1961, eight workers (26–45 years of age, mean 32 years) employed for 3.5–18 years (mean 9 years) were examined (one of these workers was also examined in 1947). Seven of the workers reported no respiratory symptoms; one worker reported a slight occasional cough. No abnormal symptoms were noted during the physical examination of seven of the workers; crepitations dispelled by cough were observed in one worker (not the same worker reporting an occasional cough). Pneumoconiosis was detected in the radiographs of seven workers. Three other workers employed for 1 month to 1 year were also examined in 1961. Two of these workers reported having slight coughs, but no abnormal findings were observed during the physical examination, and the chest radiographs were normal. The concentration of barium in the dust was not measured. Barite samples were analysed for quartz, silica, and iron content. No quartz was detected, and the total silica and total iron (as iron oxide) concentrations were 0.07–1.96% and 0.03–0.89%, respectively.

Ten of the 11 workers examined in 1961 were re-examined in 1963 (18 months later) (Doig, 1976). Two new cases of pneumoconiosis were diagnosed. Thus, 9 of 10 workers exposed to barium sulfate for 1.5– 19.5 years (mean 8.2 years) had well-marked baritosis. Three of these workers reported a slight or occasional cough, and none had dyspnoea. Among the nine workers with baritosis, three did not smoke, four smoked 1 pack/day, and two smoked >1 pack/day. In six of the seven workers with previously diagnosed baritosis, no significant changes in the degree of pneumoconiosis were observed; an increase in the number of opacities was observed in the seventh worker. Spirometric lung function tests (vital capacity, flow rate, and forced expiratory volume) were performed in five workers. For three of these workers, the results of the lung function tests were similar to predicted normal values (89–119% of predicted values). Lung function was below normal in the other two workers (70–85% of predicted values). It is questionable whether the impaired lung function was related to barium exposure. One of the two workers was an alcoholic and heavy smoker, and the other had a fibrotic right middle lung lobe that probably resulted from a childhood illness.

The barite grinding facility closed in 1964, and follow-up examinations were performed in 1966, 1969, and 1973 on five of the workers (Doig, 1976). Termination of barium exposure resulted in a decline in the profusion and density of opacities. In 1966, there was a slight clearing of opacities; by 1973, there was a marked decrease in profusion and density. No significant changes in lung function were observed during this 10-year period.

NIOSH (1982) conducted a health survey of past and present workers at the Sherwin Williams Company’s Coffeyville, Kansas, USA, facility. Work performed at the facility included grinding, blending, and mixing mineral ores. At the time of the study, four processes were in operation: "ozide process," which involved blending several grades of zinc oxide; "ozark process," which involved bagging very pure zinc oxide powder; "bayrite process," which involved grinding and mixing several grades of barium-containing ores; and "sher-tone process," which involved mixing inert clays with animal tallow. A medical evaluation was performed on 61 current workers (91% participation). Information on demographics, frequency of various symptoms occurring during the previous 2 months, chemical exposure, occupational history, and smoking history, as well as history of renal disease, allergies, and hypertension, was obtained from directed questionnaires. In addition, spot urine and blood samples and blood pressure measurements were taken. Exposures to barium, lead, cadmium, and zinc were estimated from 27 personal samples collected over a 2-day period. In the seven personal breathing-zone samples collected from the bayrite area, the levels of soluble barium ranged from 87.3 to 1920.0 mg/m³ (mean 1068.5 mg/m³), lead levels ranged from not detected to 15.0 mg/m³ (mean 12.2 mg/m³, excluding two samples in which lead was not detected), zinc levels ranged from 22.4 to 132.0 mg/m³ (mean 72 mg/m³), and all seven samples had no detectable levels of cadmium. Soluble barium was also detected in breathing-zone samples in the ozark area (10.6–1397.0 mg/m³, mean 196.1 mg/m³), ozide area (11.6–99.5 mg/m³, mean 46.8 mg/m³), and sher-tone area (114.3–167.5 mg/m³, mean 70.45 mg/m³).

Two approaches were used to analyse the results of the health survey (NIOSH, 1982). In the first approach, the workers were divided into five groups based on current job assignments. Of the 61 current workers, 14 worked in the bayrite area (mean duration 3 years). No statistically significant increases in the incidence of subjective symptoms (e.g., headache, cough, nausea) or differences in mean blood lead levels, number of workers with blood lead levels greater than 39 mg/dl, mean free erythrocyte protoporphyrin (FEP) levels, mean haematocrit levels, mean serum creatinine levels, number of workers with serum creatinine levels greater than 1.5 mg/dl, number of workers with blood urea nitrogen (BUN) levels greater than 20 mg/dl, blood pressure, or mean urine cadmium levels were observed between the different groups of workers. In the second approach, the workers were divided into seven groups based on past job assignments. One group consisted of 12 workers working in barium process areas (bayrite process and other processes no longer in operation at the facility that involved exposure to barium ores and barium carbonate) for at least 5 years; barium exposure levels were not reported for this group of workers. The results of the health survey for the barium-exposed workers were compared with results for 25 workers who stated that they had never worked in barium process areas. No statistically significant differences in mean age, number of years employed, number of current or past smokers, prevalence of subjective symptoms, mean FEP levels, mean haematocrit levels, mean urine cadmium levels, mean beta-2-microglobulin levels, or the prevalence of workers with elevated serum creatinine, BUN, or urine protein levels were observed between the two groups. The number of workers with elevated blood pressure (defined as systolic pressure >140 mmHg [>18.7 kPa] or diastolic pressure >90 mmHg [>12 kPa] or taking medication for hypertension) was significantly higher in the barium-exposed group than in the comparison group. The number of workers in the barium group with blood lead levels of >39 mg/dl was lower than in the comparison group; however, the difference was not statistically significant. Additionally, there was no sig nificant difference between mean blood lead levels in the barium-exposed workers (24 mg/dl) and the compar ison group (32 mg/dl).

The health effects associated with occupational exposure to barium during arc welding with barium-containing stick electrodes and flux-cored wires were investigated by Zschiesche et al. (1992). A group of 18 healthy welders not using barium-containing consumables in the past 10 days was divided into three groups: group A (n = 8, mean age 30.4 years) performed arc welding with barium-containing stick electrodes, group B (n = 5, mean age 43.6 years) performed arc welding with barium-containing self-shielded flux-cored wires, and group C (n = 5, mean age 32.0 years) performed arc welding with barium-containing self-shielded flux-cored wires using welding guns with built-in ventilation systems. All welders performed welding with barium-free consumables on Thursday and Friday of the first week of the study. Barium-containing consumables were used during week 2 of the study and on Monday of week 3. The subjects welded for an average of 4 h/day. The average barium concentrations in the breathing zones were 4.4 (range 0.1–22.7), 2.0 (0.3–6.0), and 0.3 (0.1–1.5) mg/m³ for groups A, B, and C, respectively. No exposure-related subjective adverse health symptoms or neurological signs were found. No significant differences between pre- and post-shift electrocardiogram, pulse rate, whole-blood pH, base excess and standard bicarbonate, or plasma concen trations of sodium, magnesium, and total and ionized calcium were observed. During week 2, decreases in plasma potassium concentrations were observed in groups A and C; the levels returned to the normal range under continuation of barium exposure and were not statistically different from levels during week 1 (no barium exposure). This drop in serum potassium levels was not observed in group B, which had a barium exposure level similar to that of group A.

Toxicity of barium to selected aquatic organisms is summarized in Table 2. A 48-h no-observed-effect level (NOEL) of 68 mg/litre was calculated for water fleas (Daphnia magna) exposed to various concentrations of barium (LeBlanc, 1980). In contrast, Biesinger & Christensen (1972) reported 48-h and 21-day LC₅₀ values of 14.5 and 13.5 mg/litre, respectively, a 16% impairment of reproduction at 5.8 mg/litre, and a 50% impairment at 8.9 mg/litre during the 21-day tests. Khangarot & Ray (1989) reported 24- and 48-h EC₅₀ (the concentration resulting in 50% immobilization) values of 52.8 and 32.0 mg/litre, respectively, for daphnids exposed to barium sulfate. A reported 48-h EC₅₀ value for developmental effects in the mussel Mytilus californianus was 0.189 mg/litre (Spangenberg & Cherr, 1996). For two aquatic amphipod species (Gammarus pulex and Echinogammarus berilloni), Vincent et al. (1986) reported 24-, 48-, 72-, and 96-h LC₅₀ values of 3980, 395, 255, and 238 mg/litre and 336, 258, 162, and 122 mg/litre, respectively, in eucalcic water; LC₅₀ values in oligocalcic water were 1260, 533, 337, and 227 mg/litre and 308, 197, 151, and 129 mg/litre, respectively. The 30-day LC₅₀ values for two species of crayfish ranged from 39 to 61 mg/litre; the 96-h values were comparable (Boutet & Chaisemartin, 1973). Heitmuller et al. (1981) reported a NOEL in the sheepshead minnow (Cyprinodon variegatus) of 500 mg/litre.

Growth of Anacystis nidulans (a cyanobacterium) in an environment containing 50 mg barium chloride/ litre was similar to that of controls. Higher concentrations of barium resulted in a concentration-related increase in growth inhibition; almost complete inhibition was reported at barium chloride concentrations >750 mg/litre (Lee & Lustigman, 1996). Wang (1986) reported a 96-h EC₅₀ (growth) of 26 mg barium/litre in duckweed (Lemna minor) in deionized water. However, in river water, barium showed no toxic effect on growth of duckweed. The lack of an adverse effect in the river water was shown to be due to precipitation of barium from the river water as sulfate. Stanley (1974) investigated the toxic effects of barium on the growth of Eurasian watermilfoil (Myriophyllum spicatum). Root weight was the most sensitive parameter measured and showed a 50% reduction, relative to controls, at a barium concentration of 41.2 mg/litre.