Assessment of Environmental Arsenic Levels in Prievidza District

www.nature.com/jea

Assessment of environmental arsenic levels in Prievidza district

T. KEEGAN,a BING HONG,a I. THORNTON,a M. FARAGO,a P. JAKUBIS,b M. JAKUBIS,b B. PESCH, c U. RANFTc M.J. NIEUWENHUIJSEN a AND THE EXPASCAN STUDY GROUP d aDepartment of Environmental Science and Technology, Imperial College of Science, Technology and Medicine, London SW7 2BP, UK bState Health Institute, Prievidza, Slovakia cMedical Institute of Environmental Hygiene, Unit of Environmental Epidemiology and Biomonitoring, Dusseldorf, Germany

A coal-burning power station in the Nitra Valley in central Slovakia annually emitted large quantities of arsenic (up to 200 tonnes) between 1953 and 1989. Since then, pollution-control measures have reduced arsenic emissions to less than 2 tonnes a year. However, the power station was still a source of airborne arsenic pollution. As part of an EU-funded study on exposure to arsenic and cancer risk in central and Eastern Europe we carried out a study of environmental levels of arsenic in the homes and gardens of residents of the district. Garden soil samples (n=210), house dust samples (n=210) and composite house dust samples (n=109) were collected and analysed using inductively coupled plasma atomic absorption spectroscopy (ICP-AES) at Imperial College. The mean arsenic content of coal and ash in samples taken from the plant was 519 g/g(n=19) and 863 g/g(n=22), respectively. The geometric mean (GM) arsenic concentration of garden soils was 26 g/g (range 8.8–139.0 g/g), for house dust 11.6 g/g (range 2.1–170 g/g) and for composite house dust 9.4 g/g (range 2.3–61.5 g/g). The correlation between the arsenic levels in soil and in house dust was 0.3 (P<0.01), in soil and composite house dust 0.4 and house dust and composite house dust 0.4 (P<0.01 for both), i.e., were moderate. Arsenic levels in both house dust and soil decreased with distance from the power station. Overall, levels in both fell by half 5 km from the point source. Weak correlations were seen between the total urinary arsenic concentrations and arsenic concentrations in composite house dust. Journal of Exposure Analysis and Environmental Epidemiology (2002) 12, 179 –185 DOI: 10.1038/sj/jea/7500216

Keywords: arsenic, skin cancer, environmental exposure, airborne, coal, Europe, Slovakia.

Introduction be increased by exposure to arsenic (Chen et al., 1986; Bates et al., 1992; Guo et al., 1994; Chiou et al., 2000). Arsenic is ubiquitous in the environment. It exists at a mean Many of these studies have been carried out in areas where concentration of 2 g/g in surface soils (Thornton and exposure to arsenic was via water, particularly in southwest Farago, 1997). In areas where the underlying geology or Bengal, Bangladesh, Taiwan and Chile. However, fewer environmental inputs produce arsenic levels can be higher. studies have addressed the possible association between In a former mining area in South West England arsenic skin cancer and exposure to arsenic from a point source. levels in soils have been measured at over 10,000 g/g Arsenic has been categorised as a class 1a human (Farago et al., 1997), though there was little evidence for carcinogen (IARC, 1987). effects on human health resulting from inhabiting those The Electrarne Novaky (ENO) power station at Novaky, areas where soil arsenic levels were high (Xu and Thornton Prievidza district, in central Slovakia has been in operation 1985). However, adverse health effects in residents of areas since the 1950s. Since that time it has emitted over 3000 where levels of arsenic are high have been observed. tonnes of arsenic, a result of burning local brown coal that Studies have suggested that incidence rates of certain has a high arsenic content. In the 1960s and 1970s the plant cancers, including those of the lung, bladder and skin, can emitted around 200 tonnes of arsenic a year, but in the 1980s the first pollution control measures were put in place and since then the levels of arsenic emitted have fallen to below 1. Address all correspondence to: Dr. M.J. Nieuwenhuijsen, Department of 2 tonnes a year. These measures included raising the height Environmental Science and Technology, Imperial College of Science, of one stack to 300 m coupled with electrostatic Technology and Medicine, London SW7 2BP, UK. Tel.: +44-20-7594- precipitators and flue-bed desulphurisation technology. 6384. Fax: +44-20-7594-6408. Most of the ash from the plant was stored in open pits Received 4 March 2002. dBencko V., Colvile R., Cordos E., Docx P., Fabianova E., Farago M., less than 5 km from the plant. A proportion of this ash was Frank P., Go¨tzl M., Grellier J., Hong B., Rames J., Rautiu R., Stevens E., used in the local manufacture of building blocks. Coal was Thornton I., Zvarova J. also burnt in homes. Since the 1980s, waste hot water from Keegan et al. Environmental arsenic levels in Prievidza District the plant has been piped to the main urban conurbations to Bratislava. It is the largest town and capital of Prievidza provide water for central heating systems. This cut the district, a region of the Trencin administrative area. The amount of coal burnt domestically. region has been a mining area for centuries; towns such as Prievidza district has had a consistently high rate of Banska Bystrica, Kremnica and Banska Stianvica are three nonmelanoma skin cancer (NMSC) since registering cancer of the original seven mining towns colonised by German incidence began in Slovakia in 1968 (Plesko et al., 2000). miners in the 14th century from where they extracted Data from the National Cancer Registry of Slovakia shows copper, gold and silver. Prievidza was principally associated that in 1996 the rates of NMSC in the district for both men with coal mining. The coal it produces from the three mines and women was the highest in the country, and rates in the around the town goes to fuel the large power station in area have been consistently high compared to other districts Novaky, 10 km to the south. since 1985. The age-adjusted rates (world standard) in 1996 were 95/100,000 and 71/100,000 for men and Garden Soil Collection women, respectively. The country rate in 1996 for NMSC Garden soil was collected from 210 households randomly was 48/100,000 for men 35/100,000 for women, and had selected from the study population of the case–control increased from a rate of 26/100,000 for both men and study. Each sample was a composite of 20 subsamples (0–5 women in 1968. The number of NMSC cases in Slovakia cm) collected with a stainless steel trowel. Where possible, increased from 587/573 (men/women) in 1968 to 1576/ the samples were collected from the householder’s own 1505 in 1996, an approximately threefold increase (Slovak garden. If vegetables were grown, soil was collected from Central Cancer Registry, 1996). However, a previous study the vegetable patch as well as from other areas of the garden. of cancers in the district found that NMSC rates were higher If the household was a flat and had no garden but an in Banska Bystrica, a nearby town, than in Novaky itself allotment nearby the sample was taken from there. If the (PHARE, 1995). household had no allotment a sample was taken either from A case–control study (EXPASCAN) was designed to the communal garden area or from a nearby amenity area, as test the possible association between the emissions for the close as possible to the entrance. In some cases (10%), none power station and the high rates of NMSC (Pesch, 2002). of these was possible and no sample was taken. Samples The routes of exposure to airborne arsenic from a point were collected in wet-strength paper craft bags and stored in source could be through inhalation, dermal absorption or the a clean, dry place until transport to Imperial College, gastrointestinal tract (Polissar et al., 1990; Farago et al., London was arranged. 2000). For this study, arsenic exposure was assessed in three House dust was collected from 210 homes. In each ways. First, total arsenic was measured in the garden soil. household two samples were collected, one was dust Also, levels of arsenic in coal and ash from the power plant collected from a measured area of carpet, the other a sample were assessed in this study, as were arsenic emissions over from the householder’s vacuum cleaner. The house dust the active lifetime of the power plant. Second, current and sample was collected using an Imperial College-adapted historical levels of airborne arsenic were modelled using an vacuum cleaner. One square meter of carpet in the main air dispersion modelling system (ADMS) based on living area was vacuumed for 2 min. The dust was collected emission data (Colvile et al., 2001). Third, arsenic uptake in a Whatman cellulose filter fitted to the inspiration pipe of from the food chain was assessed with data from a the cleaner (Watt et al., 1993). The filters were stored in preceding project, PHARE (Fabianova and Bencko, labelled, sealed plastic bags until preparation and analysis. 1995). Some routine data for local drinking water quality Where there was no carpet in the main living area another were available, but not for the past. Furthermore, urine room was sampled. (In most households the kitchen was the samples were taken from each participant and their urine main living room and was often carpeted, and this was often arsenic content measured (EXPASCAN, 2001). from where samples were taken.) In some instances, a The objective of this study was to determine and report sample too small for analysis was collected using this on the levels of arsenic in the garden soil and indoor dusts of procedure. Under these circumstances, 2 m2 of carpet were the subjects enrolled in the study. Coal burnt in the power sampled for 4 min. Duplicate samples were taken at 10% of station was also tested, and historical emissions of arsenic the addresses (1 m2 for 2 min). from the power station are presented. The composite samples were taken from the householder’s vacuum cleaner. Depending on the quantity available enough was taken to fill a 10Â4-cm plastic bag Method that was then sealed and labelled (n=109).

The Study Area Coal and Ash Prievidza is a medium-sized town (pop. 100,000) in central Since being commissioned the power station has used Slovakia, about 100 km northwest of the country’s capital a number of combinations of boilers, generators and

180 Journal of Exposure Analysis and Environmental Epidemiology (2002) 12(3) Environmental arsenic levels in Prievidza District Keegan et al.

300 Sample preparation

Soil and Garden Soil Soil and garden soil were air dried at 408C for 48 h. Each

200 sample was disaggregated using a pestle and mortar then passed through a 2-mm sieve. A portion of this sample was milled in a Tema swing mill for 2 1/2 min. This sample was bagged for analysis.

100 House dust All house dust samples were passed through a 1-mm sieve. The bag and filter used to collect the dust from a measured area of carpet had not been weighed before collection. To

Emissions (tonnes of As) 0 get a measure of the dust loading and thus, the arsenic 1967 1969 1971 1973 1975 1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 loading, 20 bags and filter combinations were weighed and Year an average taken. The full filter was weighed and the mean for an empty filter subtracted to obtain an estimate of the Figure 1. Total estimated emissions of arsenic (tonnes) from the ENO mass of dust collected from 1 m2 of carpet. plant between 1967 and 1997. Sample Digestion chimneys. Between 1967 and 1975 there were three chimneys of 100 m and one of 150 m. From 1976 to Soil From each sample, 0.3 g was weighed into a clean, dry 1979, two 10-m chimneys, one 150-m and one 300-m, test tube. In a fume cupboard, 4 ml of nitric acid was added and from 1980 one 150-m and one 300-m chimney were using an Oxford dispenser. Following this, 1 ml of in operation. These different stacks and combinations of perchloric acid was added. These tubes were then placed stacks will have led to different emissions amounts and in an aluminium heating block, itself in a fume cupboard. profiles. The total emissions by year from the power station The block’s heating cycle was set to the following are shown in Figure 1. programme: 3 h at 508C, 18 h at 1908C, 0.1 h at 1958C. Arsenic emissions were highest between 1970 and 1979. Tubes that were still emitting fumes at the end of this cycle During this period over 1700 tonnes of arsenic were continued to be heated until dry. The tubes were transferred emitted; over 50% of all arsenic emitted over the lifetime of to stainless steel racks and cooled to room temperature. the plant (3222 tonnes) was emitted during one-third of its Two milliliters of hydrochloric acid were added and the life. tubes then placed in a shallow heating block. This was The remediation technology installed at the ENO plant heated to 608C for 1 h. Afterwards the tubes were was different in each of the furnaces/boilers (ENO A and transferred to their racks. Once cool, 8 ml of deionised ENO B). ENO A, the furnace to which the 300-m water was added from an Oxford dispenser and the contents chimney was attached, relies on electrostatic precipitators of the tubes mixed using a vortex mixer. Each sample was (ESPs) to prevent particles from leaving the chimney. transferred to a polystyrene tube and capped. This fly ash collects in the ESPs. The ash was then stored The tubes were left to settle overnight then centrifuged at in a silo awaiting transport to the ash dump. ENO B has 8000 rpm for 3 min. The tubes were left in the inductively flue-bed desulphurisation technology installed. This coupled plasma atomic absorption spectroscopy (ICP- prevents the release of sulphur and heavy metals. The AES) room overnight to equilibrate to 258C. ash from this, bottom ash, was taken by open lorry to an House-dust samples were treated identically except that ash dump near the plant. Desulphurisation by-products only 1 mg of each was weighed into each test tube. This was are a mixture of bottom ash, fly ash and other waste materials. Table 1. Arsenic concentrations (g/g) in coals from the Slovak Twenty coal samples were taken, 2 from the samples power plant ENO. taken daily by the company for operational purposes, 10 n Mean Min Max from the coal store (1999 coal), 4 from the store of 1998 coal and 4 from the rail wagons in the delivery 1998 coal 2 – 338.0 499.0 QC samples 2 – 36.9 983.0 area. 1999 coal 9 690.4 292.0 1540.0 Twenty ash samples were taken, from the ESPs, ash silos, Inbound coal 4 218.3 18.5 934.0 ash dumps and from the hopper containing bottom ash and n=number of samples, Min=minimum arsenic concentration, max=maxi- waste from the desulphurisation process. mum arsenic concentration.

Journal of Exposure Analysis and Environmental Epidemiology (2002) 12(3) 181 Keegan et al. Environmental arsenic levels in Prievidza District

Table 2. Arsenic concentrations (g/g) in ash from a Slowak power materials (NBS 2709, NBS 2710, NBS 2711) were used. plant ENO. These made up 10% of the samples. Sampling and n Mean Min Max analytical variance was within acceptable levels. ESP ash 2 – 217.0 3030.0 Silo ash 2 – 1330.0 1390.0 Statistical Analysis Bottom ash 2 – 1320.0 1550.0 Desulphurisation 3 1413.3 1310.0 1530.0 All analysis was carried out on log-transformed data. by-products Ash dump I 5 226.8 124.0 329.0 Ash dump II 4 355.2 298.0 451.0 Results Desulphurisation dump 4 839.5 765.0 1010.0 n=number of samples, Min=minimum arsenic concentration, max=max- Coal from the Novaky Power Station imum arsenic concentration. Samples of the coal burnt in the power station and the ash produced by it were taken and analysed for arsenic and other because of the small sample size. The detection limit was elements. The results of the analysis of the coal are shown in not affected greatly by the smaller sample size. Table 1. The coal fed directly to the plant was designated here 1999 coal; 1998 coal was left over from the previous year Sample analysis but will be burnt. The quality control samples were the management’s own daily quality control samples. The ICP-AES Analysis inbound coal samples were taken from the rail wagons All samples were analysed using ICP-AES. This instru- parked in the delivery bays. Each of those samples was ment converts a small portion of each sample into an taken from a different wagon and each wagon came from a aerosol. The sample was atomised and the spectral different coal mine. It was not possible to infer much from emission lines are analysed by a high-resolution spectro- the range of arsenic levels in the inbound coal — this meter. The detection limit for arsenic was 0.4 g/g for soil represents 1 day’s delivery only. and 1 g/g for dust. The highest arsenic concentration was found in one sample of the 1999 coal (1540 g/g) and the mean level of Quality Control Procedures coal from the main supply was 690.4 g/g. Arsenic Strict quality control procedures were adhered to (Thomp- concentrations in both samples of coal held over from the son and Walsh, 1989). These comprised adding a selection previous year were raised, at 338 and 499 g/g, of duplicates, reference materials and reagent blanks to each respectively. The lowest concentration of arsenic in any run. Results from the analysis of the duplicates, blanks and coal sampled was found in recently delivered coal from an reference materials allow a measurement of the levels of unknown source. precision, accuracy and bias to be accounted for. Each field duplicate sample (10%) was analysed twice to Ash from the Novaky Power Station assess the level of analytical precision,. To assess accuracy, The highest mean concentration of arsenic was found in fly reference materials and reagent blanks (5–10%) were used. ash followed by bottom ash, though the actual concentra- Both in-house (HRM 10, 11) and standard reference tions were similar, at 1623 and 1360 g/g, respectively. Silo

Table 3. Arsenic concentrations (g/g) in dust and soil by distance from ENO plant. Distance (km) n GM GSD Min Max Soil <5 40 43.2 1.7 13.8 134.0 5–10 102 25.3 1.7 8.8 139.0 >10 68 21.8 1.5 9.6 58.2 Total 210 26.6 1.7 8.8 139.0 Dust <5 32 18.3 1.9 6.0 116.0 5–10 109 10.7 2.0 0.7 170.0 >10 70 10.7 2.0 0.7 112.0 Total 211 11.6 2.1 0.6 170.0 Comp dust <5 25 19.3 1.6 7.0 57.5 5–10 55 7.8 2.0 0.7 22.5 10 29 7.6 2.6 0.7 61.5 Total 109 9.4 2. 3 0.6 61.5 n=number of households, GM=geometric mean, GSD=geometric standard deviation, Min=minimum arsenic concentration, max=maximum arsenic concentration.

182 Journal of Exposure Analysis and Environmental Epidemiology (2002) 12(3) Environmental arsenic levels in Prievidza District Keegan et al. ash, which was fly ash in storage, had a mean arsenic Table 4. Correlations between arsenic concentrations in soil, dust and concentration of 1360 g/g, slightly lower than the ESP ash urine. samples, though the range of concentrations was greatly Soil Dust Total urinary Total urinary lower. Desulphurisation by-products had an arsenic con- arsenic inorganic arsenic centration similar to that of bottom ash (Table 2). Soil – 0.34** 0.21** 0.15* Ash dump I and ash dump II are the sites where ash from n 165 159 175 the silos was stored. The ash was mixed with water and House dust – 0.13 0.08 n 162 162 piped to the site. These dumps have been in operation for Composite 0.40** 0.40** 0.02 –0.05 over 20 years, though ash dump II was constructed more n 94 92 80 90 recently than ash dump I. The ash in these two dumps was n=number of samples. **P<0.01. *P<0.05. predominantly fly ash. Arsenic concentrations in these Measurement unit used for total urinary arsenic are micrograms per liter. dumps were lower than for new fly ash. It was possible that Pearson’s correlation coefficients based on log-transformed data. leaching may have removed arsenic from the ash particles into solution. Correlations Between Arsenic Levels in Garden Soils and Arsenic in Soil and House Dust House Dusts The levels of arsenic in garden soils, house dust and Levels of arsenic in house dust correlated weakly with composite house dust are shown in Table 3. Overall, levels arsenic levels in garden soils (Table 4). The relationship in garden soils were low. Levels of arsenic in house dust and between levels of arsenic in house dust and levels of arsenic composite house dust were also low, and were about half the in garden soils was weak (r=0.3). Levels of arsenic in level given for soil, at 11.6 and 9.4 g/g, respectively. The garden soil and composite house dusts was slightly stronger maximum level in house dust was higher than that for (r=0.4), the same as that between house dusts and garden soil. composite house dusts (P<0.01 for all). Arsenic concentrations in soil decreased with distance Table 4 also shows the correlations between the con- from the power station. Overall, the level of arsenic fell by centrations of arsenic in garden soil, house dusts and in about half in the distance measured. The arsenic levels fell urine. The full results for the study’s urine analysis were in steeply in the first 5 km then only by a small amount from the EXPASCAN study report (2001) and showed an ave- 5 km to the district boundary, more than 10 km from the rage (GM) urine total arsenic concentration of 6.02 g/l. plant. The level fell by 40% after 5 km but fell only a further A weak correlation was seen between arsenic concentra- 16% in the next 5 km. The range of arsenic concentrations tions in soil and total urinary arsenic, which was given as the recorded within 5 km and between 5 and 10 km from the sum of inorganic arsenic, monomethylarsonic acid (MMA), plant was similar. and dimethylarsinic acid (DMA). Soil arsenic levels also Overall, arsenic levels in dust were not as high as in correlated weakly with urinary inorganic arsenic levels. No garden soils. Mean arsenic levels within 5 km of the power correlation was observed between levels of arsenic in urine plant were 18 g/g. Arsenic concentrations fell by 40% and in house dust or composite house dust. after 5 km and did not fall further thereafter. The range of dust arsenic concentrations was slightly greater between 5 and 10 km from the plant. The standard deviations were Discussion similar for each category. The mean composite house dust arsenic content was lower than that for measured house dust Emissions of arsenic from a power plant in Novaky have samples. However, up to 5 km from the plant the levels in been high but the plant now emits less than 2 tonnes of composite house dust were higher. Beyond 5 km from the arsenic a year. Our results show that the concentration of plant the levels fell off steeply. Arsenic concentrations in arsenic in coal burnt at the Novaky power plant remains samples taken more than 5 km from the plant were less than high, with a mean level of 518 g/g. This compares to the half the levels of those taken within 5 km from the plant. normal range of arsenic in British black coal of 2–10 g/g, The difference between means of arsenic levels in soils, or US coal, where action was taken if the arsenic dusts and composite dusts within 5 km and those further concentration exceeds 5 g/g (Swaine 1994; Goldhaber than 5 km from the plant (as measured by a t-test on the log et al., 2000). The level of arsenic in ash was also raised; the transformed data) was significant (P<0.001). mean level of 1151 g/g was approximately twice that of Duplicate dust samples were taken in 24 households. the coal. The highest concentration of arsenic was found in Variation in between-house arsenic was greater than ash collected from the electrostatic precipitators fitted to the variation within house ( w =20%, b =80%) indicating 300-m chimney. Arsenic levels in ash that had been that our samples may be taken as valid measures of dumped were lower. Mean arsenic levels in dumped fly ash between-house arsenic levels. (ash dumps I and II) were 290 g/g. Dumped ash that had

Journal of Exposure Analysis and Environmental Epidemiology (2002) 12(3) 183 Keegan et al. Environmental arsenic levels in Prievidza District been produced as a result of the desulphurisation process It has been reported that the relevant exposure pathway had a mean value of 839 g/g. The ash dumps have been in for soil and dust is hand to mouth activity in children, whose operation for the lifetime of the power plant, though ash hair and urine arsenic levels have been found to be raised in dump II was newer that ash dump I. The samples were taken the vicinity of arsenic sources (Gebel et al., 1998). The age from the surface, so they represent the more recently burnt profile of this study population (mean age=64 years) would ash, yet during the residence period the ash arsenic suggest that hand to mouth activity was not a significant concentration has fallen by around 75%. exposure route. However, it has been reported that there was Overall, levels of arsenic in garden soils were low. The a direct link between soil metal levels and indoor dust levels background range for arsenic in soils in Central Europe (Fergusson and Kim, 1991) and since the levels of arsenic was between 2 and 20 g/g (Gebel et al., 1998). However, in dusts were correlated with those in soils it suggests that in the study area, the maximum arsenic concentrations in there may be an exposure pathway linking garden soil to garden soil (139 g/g) were considerably above the urinary arsenic levels. Body burden has been seen to be average. Dust in the home, in general, contained arsenic in raised in results of a locally run study. Levels of arsenic in lower concentrations than in soils. The mean arsenic hair fell with distance from the ENO power plant, though concentration for dust was 16.2 g/g, around half that overall levels were very low (Bencko, 1995). in soils. The arsenic concentration in composite house However, while distance from the chimneys had a dust, taken as an indication of longer-term exposure, was sharply measurable effect on the environmental and 10 g/g. Locally generated data shows that arsenic levels domestic levels of arsenic the same was not true of urinary in soils in Prievidza district were generally low but that arsenic concentrations. Whilst total arsenic levels fell by there were two hotspots, in the southwest of the district, 30% within 5 km of the plant, levels of inorganic arsenic in where levels exceed 70 g/g. These areas align most urine fell minimally (EXPASCAN report, 2001). This may closely with the areas around Novaky (Geologicky Ustav be explained by the low level of urinary arsenic in the study Dionyza Stura, 1995). region. A similar study found that in an area where the The best correlation existed for the relationship between arsenic content of soils and house dusts near the source arsenic in house dust and composite house dust, a were raised (267 mg/g in soil, 83 g/g in dust), the correlation equal in strength to that for the relationship population’s total urinary arsenic reached 20 mg/g between arsenic in composite house dust and garden soils. creatinine (Hwang et al., 1997); thus a 10-fold increase Since composite house dust only differed from house dust in in soil arsenic and 8-fold increase in house dust arsenic that it was a sample taken from the householder’s vacuum resulted in a 3-fold increase in total urinary arsenic. In that cleaner rather than from their carpet it was perhaps study, carried out around the Anaconda copper smelting surprising that the correlation was not stronger between plant in the United States, there was a 10-fold difference these two measures. The relationship between arsenic in between arsenic levels in ‘‘high’’ areas compared to those house dust and in garden soils was moderate, and similar to in ‘‘low’’ areas, whilst between those same areas there those found between soils and house dusts around the existed only a 2-fold difference in urine arsenic concentra- Anaconda copper smelter (Hwang et al., 1997). tions. Similarly, at the site of a mine in South West England The concentrations of arsenic in soils and house dusts fell arsenic levels in soils reached 4500 g/g, compared to a with distance from the power station. As proposed in similar reference area where levels were 37 g/g. Whilst this studies of pollution around a point source, this suggests that represents a 122-fold difference in soil arsenic concentra- the power station was a source of arsenic into the local tions, the total urinary arsenic in the contaminated area was environment (Elliot et al., 1992). The soil arsenic in the only double that in the reference area suggesting relatively vicinity of the power station was higher than in more distant little uptake (Farago et al., 2000). Since in areas where areas. Soil arsenic fell by half more than 5 km from the plant environmental levels of arsenic were high there were only but did not decrease thereafter. The actual concentration of small increases in measurable urinary arsenic, distinguish- arsenic in soils less than 5 km from the plant was 43.2 g/g, ing between exposed and unexposed individuals by the mean of samples further that 5 km from the plant was reference to their environmental arsenic levels may be 23.8 g/g. Similarly, dust and composite dust arsenic difficult. concentrations halved within 5 km of the plant but do not decrease thereafter. The decrease in arsenic levels showed a similar pattern of decay with distance of ground levels Conclusion arsenic concentrations generated by the air dispersion model (Colvile et al., 2001). Other studies have shown that decay Domestic arsenic levels were low in Prievidza district. of heavy metals around a point source is exponential, levels They were, though, higher in residences near the power fall to background levels over relatively short distances station. Levels of arsenic in both garden soils and house (Gagne and Letourneau, 1993; Pilgrim and Hughes, 1994). dusts reach near to background levels within 5 km of the

184 Journal of Exposure Analysis and Environmental Epidemiology (2002) 12(3) Environmental arsenic levels in Prievidza District Keegan et al. plant. There was a weak correlation between arsenic in soil Goldhaber M., Irwin E.R., Atkins B., et al. Environmental impact of and house dust but little with urine arsenic concentrations. elevated arsenic in Southern Appalachian Basin coals. In: Centeno J.A., Collery G., Vernet R.B., Finkleman H., Gibb J.C., Etienne J.C. The current work provides further information to assess (Eds.), Metal Ions in Biology and Medicine. John Libbey Eurotext, exposure to environmental arsenic in the population-based Paris, 2000; 38–40. case–control study of the risk of inorganic arsenic in the Guo H.R., Chen C.J., and Greene H.L. Arsenic in drinking water and development of nonmelanoma skin cancer in the study cancers: a brief descriptive review of the Taiwan studies. In: Chappell area. W.R., Abernathy C.O., Cothern R. (Eds.), Arsenic Exposure and Health. Science Technology and Letters, 1994. Hwang Y.H., Bornschein R.L., et al. Environmental arsenic exposure of children around a former copper smelter site. Environ Res 1997: 72 References (1): 72–81. IARC. Monographs of the Evaluation of Carcinogenic Risks to Humans Bates M.N., Smith A.H., and Hopenhayn Rich C. Arsenic ingestion and — Arsenic and Arsenic Compounds. WHO/IARC, Lyons, 1987. internal cancers: a review. Am J Epidemiol 1992: 135(5): 462–476. Kavanagh P. Impacts of high arsenic concentrations in South West Bencko V. Use of human hair as a biomarker in the assessment of exposure England on human health and agriculture. PhD thesis, University of to pollutants in occupational and environmental settings. Toxicology London, 2000. 1995: 101: 29–39. Kovacik M. (Ed.), Geofaktory zivotneho prostredia regionu Horna Nitra. Chen C.J., Chuang S.L., You T.M., et al. A retrospective study on Geologicky Ustav Dionyza Stura Bratislava, 1995. neoplasms of bladder, lung and liver in blackfoot disease area in Nieuwenhuijsen M.J., Rautiu R., Ranft U., Bencko V., and Cordos E. Taiwan. Br J Cancer 1986: 399–405. Exposure to arsenic and cancer risk in Central and Eastern Europe. Chiou H.Y., Chou Y.L., Teh H.W., Tseng C.H., and Chen C.J. Cancer Final report, project EXPASCAN IC15 CT98 0325. European Union, incidence and arsenic exposure among residents of Lanyang basin in Brussels, 2001. Taiwan. In: Centeno J.A., Collery G., Vernet R.B., Finkleman H., Gibb Pesch B., Ranft U., Jakubis P., Nieuwenhuijsen M.J., Hergemo¨ller A., J.C., Etienne J.C. (Eds.), Metal Ions in Biology and Medicine. John Unfried K., Jakubis M., Miskovic P., and Keegan T.J. Environmental Libbey Eurotext, Paris, 2000; 25–27. arsenic exposure from a coal-burning power plant as potential risk Colvile R.N., Stevens E.C., Nieuwenhuijsen M.J., and Keegan T.J. factor for non-melanoma skin carcinoma: results from a case-control Atmospheric dispersion modeling for assessment of exposure to study in the district of Prievidza, Slovakia. Am J Epidemiol 2002. In arsenic for epidemiological studies in the Nitra Valley, Slovakia. press. J Geophys Res 2001: 106(No. D15): 17–42. Pilgrim W., and Hughes R.N. Lead, cadmium, arsenic and Zinc in the Elliot P., and Klein J., Eschmidt I., and Westlake A.J. Use of routine data ecosystem surrounding a lead smelter. Environ Monit Assess 1994: 32: in studies of point sources of environmental pollution. In: Elliot P. and 1–20. Cuzick J. (Eds.), Geographical and Environmental Epidemiology: Plesko I., Sever G., Obsitnikova A., and Boyle P. Trends in the incidence Methods for Small Area Health Studies. OUP, Oxford, 1992. of non-melanoma skin cancer in Slovakia. Neoplasms 2000: 47: 137– Fabianova E., and Bencko V. Central European Study on the Health 142. Impact of Environmental Pollution. Project PHARE, EC/91/HEA/18. Polissar l., Lowry Coble K., Kalman D.A., Hughes J.P., van Belle G., Final report, 1995, Brussels. Covert D.S., Burbacher T.M., Bolgiano D., and Mottet N.K. Pathways Farago M.E., Kavanagh P., Thornton I., Elliott P., and Leonardi G.S. of human exposure arsenic in a community surrounding a copper Health Aspects of human exposure to arsenic in soil in South West smelter. Environ Res 1990: 53: 29–47. England. In: Arsenic Exposure and Health Effects. Chapman & Hall, Swaine D.J. Trace elements in coal and their dispersal during combustion. London, 1997. Fuel Process Technol 1994: 39: 121–137. Farago M.E., Kavanagh P., Thornton I., and Hassanienn M. Exposure to Thompson M., and Walsh J.M. Handbook of Inductively Coupled Plasma arsenic from soils and dusts in old mining and smelting areas of SW Spectroscopy. Blackie, Glasgow, 1989. England. In: Centeno J.A., Collery G., Vernet R.B., Finkleman H., Thornton I., and Farago M.E. Geochemistry of Arsenic. In: Abernathy Gibb J.C., Etienne J.C. (Eds.), Metal Ions in Biology and Medicine. C.O., Calderon R.L., Chappell W.R. (Eds.), Arsenic, Exposure and John Libbey Eurotext, Paris, 2000; 59–61. Health Effects Chapman & Hall, London, 1997. Fergusson J.E., and Kim N.D. Trace elements in street and house dusts. Watt J., Moorcroft S., Brooks K., Culbard E., and Thornton I. Metal Sci Total Environ 1991: 100: 125–150. contamination of dusts and soils in urban and rural households in Gagne D., and Letourneau G. Determining soil contamination patterns in a the United Kingdom: 1. Sampling and analytical techniques for residential district near a copper smelter. Arch Environ Health 1993: household and external dusts. Trace Subst Environ Health 1993: 17: 48: 181–183. 229–235. Gebel T., Suchenwirth R., Bolten C., and Dunkelburg H.H. Human Xu J., and Thornton I. Arsenic in garden soils and vegetable crops in biomonitoring of arsenic and antimony in case of an elevated geogenic Cornwall, England. Implications for human health. Environ Geochem exposure. Environ Health Perspect 1998: 106(1): 33–39. Health 1985: 7: 131–133.

Journal of Exposure Analysis and Environmental Epidemiology (2002) 12(3) 185