1 Introduction

Boron is an ubiquitous element in rocks, soil and water. Most of the earth’s soils have less than 10 ppm boron. Freshwaters normally range from less 0.1–1.5 ppm, with higher concentration in regions of high boron soil levels (WHO 1998). The greatest exposure to boron for most populations comes from food. Food rich in boron includes fruits, leafy vegetables, mushrooms, nuts, legumes, wine cider and beer. The daily intake of boron by humans can vary widely depending on the proportion of various food groups in the diet. Vegetarians are identified as a potential high intake group (EFSA 2004).

Water and particularly mineral water can be an important source of boron intake. The natural borate content of ground water and surface water is usually small. Natural occurring boron is present in ground water primarily as a result of leaching from rocks and soils containing borates and borosilicates. Concentration of boron in ground water throughout the world range widely from <0.3 to >100 ppm. In general, the amount of boron in fresh water depends on such factors as the geochemical nature of the drainage area, proximity to marine costal regions and imputes from industrial and municipal effluents.

According to WHO (1998) and EFSA (2004), short- and long-term of oral exposure to boron affects mainly the male reproductive system, which is a consistent target of toxicity. Furthermore, it penetrates the placenta with an embryo negative result. Until 2004, data on adverse effects of boron in humans where not adequate for establishing a tolerable upper intake level (WHO 1998). The WHO guideline value for boron in drinking water is 0.5 ppm, which is attributed to 10 % contribution from total exposure in boron. The relevant legal limit of Directive 98/83/EE is 1 ppm and the Codex Standard limit for natural mineral water is 5 ppm (1981).

EFSA (2005) defines that tolerable daily intake (TDI) for adults is 10 mg boron / kg body weight/day. With an allocation of 10 % of the TDI to drinking water (according to WHO 1998), the boron concentration must not exceed the 1 mg/kg body weight/day.

The measurement of metals and boron in drinking water at the laboratory was performed by ICP-MS. The present study demonstrates the steps undertaken for the validation of the ICP-MS method and the risk assessment carried out for boron determined in natural mineral water samples.

2 Materials and methods

2.1 Sampling and preservation

Water samples were collected in 1,000 ml PET bottles and 3 ml of (1 + 1) nitric acid solution was added to the samples for preservation. They were further stored at 4 °C and analyzed within a month.

2.2 Reagents and quality control

Suitable analytical reagents, instrument optimization / tuning solution, standards of ultra pure quality were used. Quality control system during routine analysis includes certified reference materials (NIST, LGC, NWRI, etc.), spikes, blanks, duplicates known addition samples and calibration verification standards. Moreover, the laboratory participated in appropriate proficiency testing schemes (AQUACHECK, etc.) in order to ensure the reliability and comparability of the results.

2.3 Equipment

Inductively Coupled Plasma-Mass Spectroscopy (ICP-MS): Agilent technologies 7500 ce, equipped with Micromist glass concentric nebulizer and Agilent CETAC ASX-510 auto sampler (Agilent Technologies).

2.4 Analytical procedure

The applied method which is accredited for the determination of trace metals and boron in drinking and bottled water was based on the Standard Method APHA 3125B. The specific ICP / MS measurement parameters are shown in Table 1.

Table 1 Agilent ICP / MS measuring parameters
Full size table

2.5 Method validation

The method validation included the following steps: trueness control, precision (repeatability and within-laboratory reproducibility), experiments, check of the linearity of the standard calibration curves, determination of the detection / quantitation limits and method uncertainty study. The results were satisfactory and within the accepted validation requirements. Representative calculated results of the above experiments are presented in Table 2.

Table 2 Performance characteristics of metals determination by ICP / MS
Full size table

2.6 Uncertainty estimation

The contribution of different sources of the method uncertainty was studied in the determination of the above metals at various levels. Below, an example for the procedure followed for boron uncertainty evaluation is described. The contribution of each of the following sources was studied: (a) The relative standard uncertainty of the method, determined from interlaboratory reproducibility tests (RSDRL), (b) The uncertainty from the reference calibration curve: (Uc/c), (c) The uncertainty due to the method and laboratory bias. This was determined from sixfold recovery tests: (Ubias/R), (d) The uncertainty from volumetric flasks/pipettes used for the preparation of standards solutions: (Uv/V).

The total combined uncertainty of the method was calculated using the following formula:

$$ {\text{Uc}}/{\text{C}} = \sqrt {\left( {{\text{RSD}}_{\text{RL}} } \right)^{ 2} } + \left( {{\text{Uc}}/{\text{c}}} \right)^{ 2} + \left( {{\text{U}}_{\text{bias}} /{\text{R}}} \right)^{ 2} + \left( {{\text{Uv}}/{\text{V}}} \right)^{ 2} $$

Figure 1 shows the contribution of each uncertainty source to the method uncertainty for boron determination.

Fig. 1
figure 1

Sources of method uncertainty

Full size image

3 Results and discussion

The results obtained within the implementation of the national monitoring program, indicated that metal concentrations in potable and bottled water examined, were by far below the acceptable legal limits and even below the relevant detection limits. An exception was the case of high boron levels in samples of imported carbonated natural mineral water (NMW) of the same brand. Due to the absence of upper accepted limits for boron, legal ambiguities were raised.

The Directive 2003/40/EC on natural mineral waters does not define any limits for boron, despite the fact that the deadline for setting such a limit was January 2006. The Codex Standard for Natural Mineral Waters defines the health-related limit for boron as 5 mg/L. Furthermore, Directive 98/83/EC concerning drinking water quality, defines an upper accepted limit of 1 mg/L for boron, however, excludes NMW. In addition, WHO provisional guideline value for boron (1998) in drinking water is 0.5 mg/L. Risk Assessment was performed, based on WHO recommendations (1998) and EFSA opinions (2005) which showed that water samples were considered as unsafe for human consumption. As a result, the import of the above samples was denied and the case was notified to rapid alert system for food and feed (RASFF).

The results of the samples of imported carbonated natural mineral water for boron content are shown in Table 3.

Table 3 Results of the carbonated natural mineral water
Full size table

4 Boron risk assessment

In Table 3 is shown: (a) the level found in samples of different batches from spring number 59 was approx. 13 times greater than WHO provisional guideline value (WHO 1998) for drinking water (0.5 mg/L), (b) 7 times the relevant legal limit of Directive 98/83/EE (1 mg/L), and (c) higher than the Codex Standard (1981) limit for NMW (5 mg/L). EFSA defines that boron TDI for adults is 10 mg/kg body weight/day (EFSA 2005). A 10 % of the TDI is attributed to drinking water (according to WHO 1998), so the boron concentration must not exceed the 1 mg/kg body weight/day.

Based on the above and according to article 10(a) of NMW Directive 80/777/EEC and article14 of Regulation EC No178/2002 the specific NMW was unsafe and considered to be injurious to health.

5 Conclusion

In order to satisfy both the legal aspect of the drinking water Directive 98/83/EU and Natural Mineral Water Directive 80/777/EC as well as the early response to any metal contamination, the State General Laboratory developed a validated method using ICP-MS.

According to Article 4 of the Drinking Water Directive 98/83/EC, the water intended for human consumption shall be wholesome and clean, which means that it must be free from any substances that in numbers or concentrations constitute a potential danger to human health. In the absence of legal limits for boron, it is obvious that Food Authorities look for a legal stronghold to take actions and carry out risk management especially when natural mineral waters are imported from an EU country. Furthermore, if action is taken with no legal bindings it could be considered as a barrier to free movement of goods within EU. Having in mind all the above complexities, but at the same time trying to protect Public Health, the Cyprus Risk Assessment/Risk Management Team recommended the following guidelines for natural Mineral Waters: (a) When concentration is equal or more than 5 mg of boron per l liter recommended: batches of NMW will not be allowed to be imported, and (b) when concentration is between 1.5 and 5 mg of boron per liter, appropriate labeling is used “consumption is not permitted by children and pregnant and lactating women”.

Having in mind all the above, despite the absence of maximum legal limits for substances that could pose a health risk to consumers, a national risk assessment should be carried out taking into account the measured level of these substances, national consumption figures for the general population and vulnerable groups, e. g. children, pregnant or lactating women. In the absence of legal limits, Risk Assessment can be based on WHO recommendations and EFSA opinions or other scientific opinions to minimize legal ambiguities and ensure as far as possible the protection of public health.