Assessment of Arsenic in Hair of the Inhabitants of East Croatia—Relationship to Arsenic Concentrations in Drinking Water

Abstract

The problem of elevated arsenic concentrations in water and environment is an increasing public health concern. The aim of the study was to assess the arsenic content in human hair in selected areas of eastern Croatia and to compare them with measured values after installation of a new water supply system. The hair samples were taken in the areas of wider Osijek and Vinkovci area and analyzed using the ICP–MS method. These data were also compared with data for Vinkovci previously published in 2004. Depending on the investigated area, the median concentrations ranged from 0.02 to 0.9 µg g⁻¹, whereby this last value exceeded the upper range of the reference value (0.319 µg g⁻¹). The arsenic concentrations from the Našice, Osijek and Vinkovci areas were within or slightly above the maximum allowed reference range. The highest median values in hair samples were detected in Čepin, with arsenic-contaminated potable water, while in areas where the water source was changed, the values were significantly lower. The results add to the conclusion that there has been significant reduction in hair arsenic concentrations in the population that was given access to clean, uncontaminated water from other regional sources.

1. Introduction

The presence of arsenic (As) in drinking water is an important public health problem in the world since arsenic is considered a dangerous geogenic contaminant of today. Groundwater (but also surface water) can be contaminated with arsenic due to the natural sources and/or due to the anthropogenic activities. Natural sources are a result of the geological composition of soil and sediments, as well as the hydrogeological characteristics of the area. Environmental arsenic pollution, as a consequence of anthropogenic activities, includes: ore smelting, emissions of fossil fuel combustion gases, timber production, pharmaceutical and glass industries, use of arsenic-based pesticides and herbicides and the disposal of industrial chemical waste. More recently, arsenic has also been used as an antibiotic in poultry farming, and through feces and degradation by nitrates and organic matter, also present in the feces, it enters the environment in the form of arsenates and arsenites [1,2,3,4].

Elevated As in drinking water has been reported for many areas of the world including Chile, Mexico, Argentina, China, Taiwan, United States, Canada, India and Bangladesh [4,5,6,7].

In Europe, arsenic-contaminated groundwater is found in different areas such as Central Europe [8,9], Spain, Slovakia, Switzerland and the UK as well [10,11,12]. In Eastern Europe it has been reported in the alluvial aquifers of the Pannonian Basin: western Hungary, northeastern Romania, northern Serbian region of Vojvodina; in the alluvial deposits of the river Axios and in the karst aquifers in Kalikratia, northern Greece; Bosnia and Northern Macedonia. Arsenic-contaminated water is used for public water supply, but there are exceptions in some places, such as Zrenjanin in Serbia, where measures prohibiting the use of tap water have been enforced since 2004, except as “technical” water for industrial use, due to the elevated arsenic levels [13,14,15].

The presence of arsenic in the water supply systems of eastern Croatia, especially in the Vinkovci and Osijek regions, provoked the interest of the general public, and due to the high values of As published in the media, it raised concerns among the local population [16,17,18,19]. In eastern Croatia, the presence of arsenic in water is the result of hydrogeochemical properties of the sediment in aquifers belonging to the Drava river basin and partly to the Danube region [20]. According to research by Ujević et al., drinking water sources from the Sava river basin have up to 10 times lower values of arsenic than water coming from the Drava or Danube river basins. In the inner parts of the Drava and Danube depressions, As concentrations range up to 70 and 220 µg L⁻¹ [21]. It should be emphasized that according to some authors, high values of arsenic in the soil can be found in gardens in these areas, possibly as a result of irrigation with tap water, and in dandelion and cabbage as phytoindicator plants, which are also part of the food chain [18,19].

The first two reports on arsenic in groundwater of Croatia were made in 2002 and 2004, and it raised serious concern in wider public. Ćavar et al. [16] in their research measured higher values of arsenic in hair in Osijek and Vinkovci ranging from 1.74 to 4.31 µg g⁻¹ in subjects, while drinking water in the area of Vinkovci had arsenic concentrations ranging from 17.60 to 611 µg L⁻¹. At that time, the national legislation allowed a maximum As concentration of 50 µg L⁻¹, while the EU regulation allowed 10 µg L⁻¹. This exemption was temporarily allowed by the European Union, until a new system of water supply was introduced [22]. These measurements also showed that in Čepin (near Osijek) the maximum water concentration in the water exceeded the maximum permitted level set by the EU (i.e., 10 μg L⁻¹) more than 16 and 25 times. These results led to the closure of the local water system belonging to the Drava River basin in Čepin near Osijek and immediate activation of the regional Sava River-based water supply in Sikirevci. Vinkovci area is now completely supplied from the regional water system in Sikirevci and the wider region [23,24], and Osijek is expected to start using filtered water in 2021. According to its authorizations, the Ministry of Health of the Republic of Croatia has raised the permitted values of arsenic in water to 50 µg L⁻¹ by July 2019 (despite being an EU member state with maximum allowed concentrations of 10 µg L⁻¹), but also banned the use of drinking water from local waterworks in Čepin and the wider Vinkovci area due to measured values of 100–220 µg L⁻¹. This caused great concern and even protests by locals in these settlements [16].

Arsenic has been described as carcinogenic by World Health Organization’s International Agency for Research on Cancer (WHO IARC) for lung, bladder and skin cancer and potentially linked to kidney, liver and prostate cancer [25,26]. Arsenic has also been linked to cardiovascular diseases, reproductive problems and skin and neurologic disorders [27,28,29].

Determination of a correlation between arsenic in hair and in water reflects the relationship between exposure, absorption and excretion. Although there are additional measurements of arsenic metabolites available, such as monomethylarsonic acid and dimethylarsinic acid, there is still potential for the use of arsenic in hair as a potential screening tool for research and clinical use as described by Katz [30].

This is the first research of this kind, in this area and among the population in east Croatia, after the introduction of new aqueduct. We hypothesize that the introduction of drinking water with low arsenic level will result in lowering of arsenic concentrations in Croatian population. Hence, this study aimed to investigate the arsenic values in hair of the population of east Croatia from areas that have elevated arsenic concentrations in drinking water and to determine whether reduction in arsenic in drinking water will result in lowering arsenic concentrations in the hair in areas previously endangered by elevated arsenic in drinking water.

2. Materials and Methods

2.1. Study Area and Participant Recruitment

Sampling was performed successively over a 2-month period in 2015 at locations within the Osijek Baranja County—the local municipalities of Vladislavci (with adjacent Dopsin, Hrastin), Čepin, Osijek and Našice, and Vinkovci in the Vukovar Srijem County (Figure 1). The locations Vladislavci, which include smaller villages of Dopsin and Hrastin and Čepin are small villages without industry of any kind, where the population mainly subsists from farming and agriculture, with low traffic in the area located approximately 20 km southwestern of Osijek. Osijek is the industrial center of eastern Croatia with agricultural, metal, chemical and food industries. Although the town of Našice is part of the Osijek Baranya County, their local water system does not come from the Drava basin and their water has never had elevated As values. Vinkovci is the largest town in the Vukovar Srijem County with food, textile and building industries, and heavy traffic, since it is part of the 10th European railway and road corridor. The water in use belongs to local water sources, whether they are local wells or small aqueducts. These areas belong to the Drava basin, as well as the city of Osijek. The area of investigation is known for having published increased values of arsenic in drinking water and is also part of the Drava and Danube River basins [16,20,21].

The examinees in the study were 311 adults from the surveyed area with 289 from Osijek–Baranja County and 26 from the area of Vinkovci, Vukovar-Srijem County. All of the examinees were non-occupationally exposed to arsenic, their hair was not dyed and their residence was in the investigated area in the past 25 years (since Croatia’s homeland war 1991–1996). An interview was implemented with a targeted questionnaire about basic demographic indicators and habits as well as potential exposure to metals and metalloids in pre-war, wartime and post-war periods for each of the subjects. In addition, sex, age, area of residence, smoking, alcohol consumption and origin of drinking water used were also noted. The subjects answered a questionnaire with 14 questions, of which 12 were offered with multiple choice answers and 2 with open-ended answers (age and previous diseases). Possible exposure to war metals from past homeland war, domicile exposure (living near a possible source), possible occupational exposure to metals, pesticides and fossil fuels were also noted. Since the study is a part of the project “Research of long-term consequences of the war on public health” (Ministry of Science, Republic of Croatia), participation in the war (civilian, soldier, wounded, etc.) activities were also noted. Informed consent was signed by all of the examinees and ethical approval was obtained by Faculty of Medicine in Osijek.

2.2. Hair Sampling

Hair samples were taken from 311 subjects from the occiput (approx. 1 cm width and 3 cm length) with stainless steel scissors and stored in polyethylene bags. Hair samples were washed in distilled water, soaked 1 hour in acetone, rinsed with deionized water 10 times, and dried in the air on filter paper for 24 h. For each 0.1 g of sample, 1 mL of 65% nitric acid was added. After 2 hours of soaking, the samples were placed in a microwave oven (Ethos D Microwave Labstation, Milestone, Brondby, 1996), heated, cooled, diluted to a volume of 12 mL and transferred to an automatic sampler cuvette (AS 93 plus, Perkin Elmer, 2008). The settings of the microwave oven for incineration of samples were: 5 min at 250 W, 5 min at 400 W, 5 min at 500 W and 10 min of ventilation. Mass spectrometry analysis was performed using inductive coupled plasma (ICP-MS; ELAN DRC-e, Perkin Elmer, Waltham, MA, USA). The settings of the ICP-MS were as follows: voltage power (RF), 1050 W; argon > 99.99% (Messer, Sulzbach, Germany). The phases and speeds of flow through the torch were: (1) the phase of plasma flow between the external and middle columns, speed 15.00 L/min; (2) the phase of auxiliary flow of gas, speed 1.20 L/min; (3) the phase of argon gas nebulizer flow through the induction column, speed 0.88 L/min. All samples were analyzed by mass spectrometry, with inductive coupled plasma (ICP-MS ELAN DRC-e, Perkin Elmer, Waltham, MA, USA). The instrument was recalibrated after every 12th sample, using an external standard (71-Element Group Multi Element Standard Solution, Inorganic Ventures, Christiansburg, VA, USA), with the use of internal standards with elements Y, In, Tb and Bi (Inorganic Ventures, USA). Internal calibration (an international laboratory audit) was conducted in cooperation with IFA Tulin (Department of the University of Natural Resources and Applied Life Sciences, Vienna, in cooperation with the Vienna University of Technology and the University of Veterinary Medicine). Analytical methods were validated with standard reference materials and standard samples. Polyatomic interferences with elements Fe, As, Cr were removed in the dynamic reaction chamber of the instrument, with reactive methane gas (CH₄). After that, analysis by mass spectrometry with inductive coupled plasma followed (ICP-MS; ELAN DRC-e, Perkin Elmer, Waltham, MA, USA) [31]. Analytical methods were validated with standard reference materials and standard samples.

All steps of sample preparation and analysis were completed in a laboratory with standard heating, ventilation, and air conditioning (HVAC) systems combined with high-efficiency particulate air (HEPA) filters. Calibration curve standards prepared in 1% HNO₃ were used for quantification Calibration curve standards prepared in 1% HNO₃ were used for quantification, while limit of detection (LOD) and quantification (LOQ) were calculated as mean plus 3 times (LOD) and 10 times (LOQ) of the blanks’ standard deviation. The LOD were 0.002–0.006 µg g⁻¹ (LOQ 0.006–0.02 µg g⁻¹). Commercially available reference materials: human hair IAEA-086 (IAEA, Vienna, Austria), human hair NIES No. 13 (NIES, Tsukuba, Ibaraki, Japan), and bovine liver NIST SRM 1577a were used for checking the analytical accuracy of the measurement. Commercially available reference materials: human hair IAEA-086 (IAEA, Vienna, Austria), human hair NIES No. 13 (NIES, Tsukuba, Ibaraki, Japan), and bovine liver NIST SRM 1577a (Sigma-Aldrich, Darmstadt, Germany) were used for checking the analytical accuracy of the measurement. In this project, (beside As) 27 elements were analyzed, and all were covered by at least one of the selected reference materials.

2.3. Water Sampling and Analysis

Water samples were taken from local water supply systems, from local wells or from water taps in houses or public buildings in the same week as the hair sampling was performed.

Water samples were stored in polyethylene water sampling bottles of 250 mL, which were previously prepared by soaking for 24 h in 10% nitric acid at room temperature. The packaging was stored for transport in polyethylene bags. During sampling, drinking water samples were acidified by the addition of 10 mL/L concentrated HNO₃ solution (65%) and stored in polyethylene bags against external contamination. On the same day, the samples were transported to the laboratory, where they were stored at −30 °C until analysis. Water samples, after defrosting at room temperature for 24 h, were measured directly on ICP-MS, without further processing of the samples. Samples were analyzed by inductively coupled plasma mass spectrometry (ICP-MS; ELAN DRCs, Perkin Elmer, Waltham, MA, USA).

Operating conditions on ICP-MS were: voltage (RF) −1050 W; argon > 99.99% (Messer, Sulzbach, Germany). Phases and argon flow rates through torch were:

• 1st phase of flow—plasma flow between outer and middle column—speed 15.00 L/min.
• 2nd phase of flow—the auxiliary gas flow—speed 1.20 L/min.
• 3rd phase of argon gas flow—cooling (nebulizer) flow through the induction column—speed 0.88 L/min.

The instrument was recalibrated after each 12th sample, by an external standard (71-Element Group Multi Element Standard Solution, Inorganic Ventures, Christiansburg, VA, USA), using internal standards with elements Y, In, Tb and Bi (Inorganic Ventures, Christiansburg, VA, USA). Intercalibration (international laboratory testing) was conducted in collaboration with IFA Tulin (Department of the University of Natural Resources and Applied Life Sciences, Vienna, in collaboration with the Vienna University of Technology and the University of Veterinary Medicine). The analytical method was validated with standard reference material Certified Reference Material ICP multi element.

Standard solution X for surface water testing (CertiPUR^®, Merck, Darmstadt, Germany) accuracy levels of As in water samples (measured value 52.2 µg L⁻¹; certified value 53.0 µg L⁻¹, analytical recovery value was 98.49%; LOD was 0.04 µL⁻¹). Polyatomic interferences with elements Fe, As, Cr, were removed in the dynamic reaction chamber of the instrument using reactive methane gas (CH₄).

One exception was for the location of the regional aqueduct in Sikirevci, which supplies the Vinkovci area, which is run by the local authority, where the sample was provided to us and the result was given as a single measurement, since it is taken on an annual basis and published and reported annually by the Public Health Institute. Determination of total arsenic concentrations was performed by the anodic stripping voltammetry technique (ASV). Arsenic concentrations were determined in 10 mL of the samples, acidified with 10 mL of 30% HCl. An electrochemical instrument (Metrohm 757 VA Computrace, Herisau, Switzerland) with a laterally placed rotating gold electrode was used for measurement. The test was performed daily with standards of 10 and 200 μg L⁻¹ (Carlo Erba, Cornaredo, MI, Italy) [32].

2.4. Data Analysis

The Kruskal–Wallis test of differences between independent data groups was used. The method was chosen because it is nonparametric, does not depend on the normality of the distribution, and with the significance of the overall differences, it separately observes the differences between data groups, which greatly facilitates interpretation. The level of statistical significance of p < 0.05 was used. The (dis)similarities were examined by means of cluster analysis.

3. Results and Discussion

Arsenic concentrations in water and hair of respondents are presented in

Table 1. Arsenic values in drinking water (µg L⁻¹) and hair (µg g⁻¹). Respondents who used only water are marked with unexposed, while those who were potentially occupationally exposed to external influences (fossil fuels, industry, fertilizers, etc.) are marked with exposed.

Location	N	Water (µg L−1) M (25–75%)	N	Hair (m + f) (µg g−1) M (25–75%)	N	Hair (m) (µg g−1) M (25–75%)	N	Hair (f) (µg g−1) M (25–75%)	N	Hair (exp.) (µg g−1) M (25–75%)	N	Hair (unexp.) (µg g−1) M (25–75%)
Vladislavci	16	1.3 (0.7–13)	90	0.4 (0.2–0.6)	40	0.3 (0.2–0.7)	50	0.3 (0.2–0.4)	49	0.3 (0.1–0.7)	41	0.3 (0.1–0.4)
Čepin	12	191.2 (180–209)	51	0.9 (0.5–1.3)	38	0.9 (0.6–1.4)	13	1.1 (0.4–1.6)	28	0.9 (0.4–1.3)	23	0.7 (0.6–1.1)
Našice	9	0.08 (0.05–0.09)	80	0.05 (0.03–0.09)	31	0.05 (0.03–0.1)	49	0.05 (0.02–0.07)	31	0.04 (0.02–0.07)	49	0.07 (0.03–0.1)
Osijek	12	24.2 (16.3–29)	64	0.3 (0.1–0.5)	56	0.4 (0.2–0.5)	10	0.14 (0.1–0.2)	26	0.4 (0.2–0.6)	38	0.2 (0.1–0.5)
Vinkovci	SM	1.7 *	26	0.02 (0.01–0.03)	17	0.02 (0.01–0.03)	9	0.004 (0.001–0.01)	10	0.02 (0.01–0.03)	16	0.02 (0.01–0.03)

f—women; m—men; exp.—exposed, unexp.—unexposed; M—median; 25–75%—percentile; N—number of samples. Vladislavci—private shallow water wells; Čepin—local water supply system; Našice—local water supply system; Osijek—city aqueduct, SM-Vinkovci—single measurement, *—supplied from Sikirevci regional water supply system, for the official state annual arsenic monitoring.

. Since the data were not normally distributed median and interquartile ranges (25% to 75%) were given.

Basic demographic characteristics of the examinees and source of water used for drinking are presented in

Table 2. Basic demographic characteristics of examinees and source of water used for drinking.

Locations	Vladislavci	Čepin	Našice	Osijek	Vinkovci
Male respondents	40	38	31	51	17
Female respondents	50	13	49	13	9
Occupational exposure	50% male 49% female	60% male 15% female	45%male 36%female	35% male 46% female	51% male 11% female
Available watersource	private shallow water wells	local water supply system	local water supply	city aqueduct	supplied from Sikirevci regional water supply system *
Source of drinking water used by the respondents	80% water well 20% bottled water	88% local water supply system 12% bottled water	100% local water supply system	1.3% local private wells 98.7% city aquaduct	11%city aquaeduct and bottled water 89% city aquaeduct

.

Determination of a correlation between arsenic in hair and in water shows the relationship between exposure and absorption. The results from Figure 2, Figure 3 and Figure 4 show a strong positive correlation between the concentrations of As in the hair of male and female respondents and its concentrations in drinking water, where higher As concentrations in water led to higher concentrations in the hair samples (r_water/male = 0.92; r_water/female = 0.96). In Čepin area, the median arsenic concentration found in the hair ranged from all: 0.9 μg g⁻¹, male: 0.9 μg g⁻¹ to 1.1 μg g⁻¹ (female) exceeding the reference range of 0.034–0.319 μg g⁻¹ by three times. In areas with moderately elevated concentrations of arsenic in water samples (Osijek: 24.2 μg L⁻¹), median values of arsenic in hair (all 0.3 μg g⁻¹, male: 0.4 μg g⁻¹, female: 0.14 μg g⁻¹) samples ranged between the values of control (Vinkovci, Našice) and exposed areas (Čepin), and according the reference range this level is considered as a normal. A recent study showed that arsenic concentration is significantly correlated with well depth and this result suggests that the highest As concentrations are associated with a contribution of water from deeper aquifers (>70 m depth) [20,21]. According to previous investigations in Osijek–Baranja County, private shallow wells (8–13 m depth) which contain mainly Holocene sediments, are not affected by As [20].

For example, in Vladislavci the maximum water concentration in the water (old shallow wells) exceeded the maximum permitted level set by the EU (i.e., 10 μg L⁻¹) only in one case (160 µg L⁻¹). In the remaining private shallow wells As concentrations were very low (median As = 1.3 µg L⁻¹). Out of a total of 90 respondents, 80% of them stated that they consume well water, while the rest stated that they consume a combination of well water and bottled water.

In the Našice area, where the median measured As concentration in water was 0.08 μg L⁻¹, the median As concentration in hair samples was very slightly (0.050 µg g⁻¹) above the lower limit of the reference range (0.034–0.319 μg g⁻¹).

Before 2010 Vinkovci and Čepin had a similar problem. In local wells, according to Ćavar et al.’s research published in 2005, the arsenic level in water and hair samples were as high as 611.89 µg L⁻¹, and 4.31 µg g⁻¹. This prompted a wider and immediate reaction, so a new water system has reconnected the Vinkovci water supply system with wells 60 km away near Sikirevci, in the Sava River basin, an area low in arsenic. The Sikirevci regional water supply system delivers water with a low As content of 1.7 µg L⁻¹ from a depth of 75 m [16].

In the subjects from the Vinkovci area, the measured values of As in the hair was low (from 0.004 (female) to 0.02 μg g⁻¹ (male)), which can be explained by the long term use of arsenic-poor water in the Vinkovci area (Table 1).

The result of the Kruskal–Wallis test indicated the existence of statistically significant differences between the concentrations of arsenic in the hair of the subjects of Vinkovci and Našice and the hair of the subjects Čepin, Vladislavci and Osijek (p = 0.00). No statistically significant differences were found between the arsenic concentrations in the hair of Vinkovci subjects and the arsenic concentrations in the hair of Našice residents (p = 0.784193), both regions that use low arsenic value drinking water.

When these values are compared with the ones published by Rodushkin and Axelsson, then it is clear that examinees from all areas except of Vinkovci and Našice had values several times or slightly above referral values i.e., range 0.034–0.319 μg g⁻¹ [33].

The values of arsenic in the hair of the respondents from the settlement of Čepin are the highest among the examined settlements and significantly higher in relation to the other examined settlements.

The values of arsenic in the hair of the respondents from the settlement of Čepin were the highest among the examined settlements and significantly higher in relation to the other examined settlements. There were no significant differences (p = 0.8) in hair concentrations of the arsenic in hair between potentially occupationally exposed and unexposed subjects (either exposed to war contamination or occupationally exposed to arsenic). Thus, it demonstrates that, in this case, arsenic concentrations in water (not the other, external sources of arsenic exposure) have had the largest influence on measured arsenic concentrations in hair.

In 2001, the Agency for Toxic Substances and Disease Registry (ATSDR) panel identified the factors that limit the interpretation of even the most accurate, reliable, and reproducible laboratory results. These limitations were due to the lack of a reference range in which to frame the results, difficulties in differentiating exogenous from endogenous depositions, and problems with the extent of environmental contaminant incorporated into the hair. Moreover, other problems were identified such as the lack of correlation between levels in the hair, blood, and other target tissues as well as the lack of epidemiological data linking the levels of specific substance in the hair with adverse health effects [31].

Although there was no agreement on a well-defined international standard for the “normal” level of arsenic in human hair, from time to time, national and international agencies have prepared reference materials. [34]. However, the total arsenic concentrations in these materials could reflect the “normal values” only for the regions from which the hair was collected; i.e., China 0.282 µg g⁻¹, Japan 0.10µg g⁻¹. Belgium 0.044 µg g⁻¹, Germany 0.13 to 3.71 µg g⁻¹. Despite the open issues in the use of hair analysis for evaluation of arsenic intoxication, Katz states that arsenic levels above 1 or 2 µg g⁻¹ are indicative of arsenic exposure, but says that this should be followed up with clinical signs or symptoms [30]. Given the maximum allowable concentration of arsenic in water for human consumption of 10 µg L⁻¹, which was agreed on the basis of epidemiological studies conducted in the population exposed to far higher concentrations of arsenic than those recorded in Croatia, it is considered that the population is exposed to arsenic in concentrations higher than allowed in the water they use in the long-term, and is, therefore, at risk of developing some of the described diseases that arsenic can potentially cause [26].

There have been numerous episodes of arsenic poisoning around the world, with exposure coming from natural groundwater contamination, industrial sources or contamination from food and beverages. Some of the most infamous were the incidents on the south-west coast of Taiwan [35,36] with cases of Blackfoot disease (BFD, hyperkeratosis) caused by As in the water well; Antofagasta, Chile with high arsenic in drinking water first affecting children and West Bengal, India, exposing over 6 million to arsenic from groundwater [4,37,38,39]. Similar incidents were reported in Mexico, Argentina, west Oregon, California and Utah in USA, as well as in Canada [4]. From the perspective of this study, Hungarian problem with arsenic in drinking water measured in water wells is important, since investigated area in Croatia and neighboring Hungary share the River Danube and have a similar geological and geographical background. The amount of arsenic present in the well water was in the range of 60 to 4000 µg/L. A few thousand people showed symptoms of arsenic poisoning such as melanosis, hyperkeratosis, skin cancer, internal cancer, bronchitis, gastroenteritis and hematologic abnormalities. Recently, concentrations of arsenic above 50 µg L⁻¹ have been identified in groundwaters from alluvial sediments associated with the River Danube in the southern part of the Great Hungarian Plain [40,41]. Concentrations up to 150 µg L⁻¹ (average 32 µg L⁻¹, 85 samples) were found by Varsanyi et al. The groundwaters have highest arsenic concentrations in the lowest parts of the basin, where the sediment is fine-grained [42]. A few thousand people were affected and several symptoms of arsenic poisoning such are melanosis, hyperkeratosis, skin cancer, internal cancer, bronchitis, gastroenteritis, hematologic abnormalities were observed [4].

The relationship of inorganic arsenic in the drinking water to various cancer is well documented at high exposure levels, mainly at levels >200 µg L⁻¹. The risk of lower exposures is less clear and, from an epidemiologic perspective, still needs to be established. Epidemiologic evidence has suggested a relationship between arsenic exposure and cancers of the bladder, kidney, liver, and lungs [35,43,44]. In the 1980s, US Environmental Protection Agency (USEPA) reviewed carcinogenicity of inorganic arsenic and developed a cancer slope factor for risk assessment of ingested arsenic based on linear extrapolation of skin cancer at high doses (mid-points of these groups were 170, 470 and 800 µg L⁻¹) in Taiwan with elevated arsenic levels in well waters. In 2001 USEPA decided on maximum contamination level of 10µg/L, but in 2010, a revised evaluation proposed even lower dose-response slope for cancer risk [45,46].

Lynch et al. [46,47] have calculated a lower risk assessment than the USEPA based on meta–regression analysis of epidemiological studies on lung and bladder cancer. Tsuji et al. [36] performed research on dose response for assessing the cancer risk of arsenic in drinking water on in vitro and animal models and determined a drinking water threshold of 65µg L⁻¹ to be associated with in vitro biological effects that are associated with urinary bladder, lung and skin cancer. They concluded, based on mechanistic, in vitro and in vivo research, and relevant epidemiological studies with individual-level data, that a threshold level for arsenic in drinking water is estimated to be around 100 µg L⁻¹ with strong evidence that the threshold is between 50 and 150 µg L⁻¹ [34].

Lamm et al. conducted a large ecological study on low levels of arsenic (<50 µg L⁻¹) in drinking water and lung cancer for US counties finding negative association for lung cancer on such low doses [44].

Chronic exposure to arsenic is also linked to cardiovascular diseases, noncirrhotic portal hypertension, stroke, peripheral vascular disease, skin conditions and perinatal morbidity such are low birth weight and infant mortality [48,49,50,51,52]

Wu et al. estimated 0.3–0.6% of cardiovascular mortality in India is due to exposure to the groundwater arsenic. They concluded that between 58 and 64 million people in India are exposed to arsenic in water exceeding 10 µg/L [53]. Due to the possibility of oral ingestion not only through water, but also through diet with crops irrigated by contaminated water from the wells, the contribution of arsenic exposure through food chain should not be ignored, in particular through rice. Jayasumana et al. [54] reported that phosphate fertilizers are the main sources of arsenic and are linked with chronic kidney disease in Sri Lanka. Bofetta and Borron [55] performed a meta-analysis of low-level exposure to arsenic in drinking water and risk of lung and bladder cancer and did not find evidence of increased risk of bladder and lung cancer in concentration of arsenic up to 150 µg L⁻¹. A review of the Health Effects of Arsenic Longitudinal Study (HEALS) data on 20,000 subjects in Bangladesh with elevated well water arsenic exposure found evidence for increased pre-malignant skin lesions, high blood pressure, neurological dysfunction, and all-cause mortality [56].

Moon et al. conducted research on chronic arsenic exposure and cardiovascular disease at levels above 50 µg L⁻¹ with elevated risk for cardiovascular disease, coronary heart disease, stroke and peripheral arterial disease. However, at lower doses, results were inconclusive [57,58]. Xu et al., conducted a systematic meta-review on cardiovascular disease, carotid atherosclerotic disease and coronary heart disease. They found a significant positive association even at concentrations of 10 µg L⁻¹, suggesting a further lowering of the guideline value for arsenic in drinking water. [59].

However, according to the literature, the most documented severe consequences for the population have been reported at values above 600 µg L⁻¹ [26,60,61]. It should be emphasized that acute consequences of arsenic poisoning (i.e., “Blackfoot disease” etc.), at measured values in the population of eastern Croatia, have never been reported and can hardly be expected, due to increased awareness of As presence in local waterways. Taking into account all of the above, and given the maximum allowable concentration of arsenic in water for human consumption of 10 µg L⁻¹, (and especially, areas with As above 50 µg L⁻¹) the population of certain areas in eastern Croatia could be considered as long-term exposed to arsenic and is therefore still a possible long-term risk for cancer and associated cardiovascular diseases. Capak in his thesis, estimated that 74,177 men and 83,295 women were exposed to increased arsenic in water at the area of Osijek Baranya County alone. This would be accounted if EU legislation of maximum allowed concentrations of As in drinking water of 10 µg/L⁻¹ would apply [26].

Smith et al. have explored lung, bladder and kidney cancer mortality in northern Chile 40 years after reduction of arsenic in drinking water concluding that there is increased risk of manifestation after exposure reduction suggesting a possible long cancer latency for arsenic as human carcinogen [62]. Saint- Jacques et al. completed comprehensive research on arsenic in drinking water, suggesting that with exposures to 50 μg/L, there is an 83% probability for elevated incidence of bladder cancer, and a 74% probability for elevated mortality. For both bladder and kidney cancers, mortality rates at 150 ug/L were about 30% greater than those at 10 μg/L. Their meta-analyses suggest that exposure to 10 μg/L of arsenic in drinking water may double the risk of bladder cancer, or at the very least, increase it by about 40% [63].

The Croatian government has decided to introduce new a regional water supplying system from Sava Basin to insure availability of drinking water with low arsenic. Other possibilities for arsenic removal include oxidation techniques, coagulation–flocculation, membrane technologies, adsorption and ion changes (using variety of sorbents such are char, activated carbon, red mud etc.) or application of nanoparticles [64,65].

4. Conclusions

Arsenic is one of the most significant geocontaminants of today, with numerous acute and chronic (including carcinogenic) effects, affecting all body systems. The occurrence of arsenic in elevated concentrations in the local water supply systems of eastern Croatia is not an isolated phenomenon, because it is a systemic hydrogeochemical phenomenon that is characteristic of the wider area of the Danube and Drava Rivers. The hair from the subjects from the Čepin, Vladislavci and Osijek contained significantly higher arsenic values than the reference values published in the literature. This could be due to the consumption of drinking water from local waterworks rich in arsenic. The values of arsenic measured in the hair of the subjects in the area of Vinkovci were lower than those previously published in 2005, and are lower than other values measured in subjects from the settlements belonging to the Drava and Danube basin. This could be attributed to the introduction of a new aqueduct and changing the water supply for the city of Vinkovci and the surrounding area, making available water of excellent quality with a very low heavy metal content.

The most important measure to be taken as soon as possible is to stop the exposure of the population, i.e., to enable the population to be supplied with healthy water, as has been conducted in many other places, by providing an alternative supply, filtering, or a connection to wider regional waterworks. Studies of this kind should serve as an incentive for wider action, since arsenic in water threatens over 1 million people in the affected Balkan area.

Finally, it should be emphasized that since this is the first systematic research of this kind in this part of Croatia, additional extensive public health research is needed in order to put the data obtained in this research into a wider epidemiologic framework.