Potential health impacts of consuming desalinated bottled water

As water scarcity increases globally, many countries are forced to supplement existing drinking water resources with desalinated water to meet rising demands. Approximately 75 million people worldwide rely on desalinated water with this population increasing as fresh water shortage is realized throughout the globe (Avni et al. 2013). Qatar, like many other arid nations, is completely dependent on seawater desalination technologies for potable drinking water, including bottled water (Darwish et al. 2013). This dependency creates a condition of high vulnerability to compromises in drinking water quality and quantity due in part to inadequate storage capabilities and the risk of source water contamination.

Currently, bottled water is the primary source of drinking water in Qatar and other Gulf states, due predominantly to the public perception of increased quality in comparison to municipal tap water (Wait 2008). This opinion is prevalent despite the fact that bottled water has been shown to contain higher levels of some contaminants due to inadequate regulation and monitoring, pollutant content of source water and packaging and storage conditions that give rise to trace element contamination (Shotyk & Krachler 2007a, 2007b; Westerhoff et al. 2008; Guler & Alpaslan 2009; Krachler & Shotyk 2009). Bottled water consumption is an important route of trace element exposure as the continued intake of heavy metals in drinking water, even at low concentrations, has been associated with various cancers, neurotoxic effects, central nervous system disorders, cardiovascular system impairments and reproductive and genetic effects (Virtanen et al. 2007; Mansouri & Cauli 2009; Martins et al. 2009; Whitfield et al. 2010; Olawoyin et al. 2012).

The trace element and micronutrient content of bottled water is highly varied, as highlighted in the meta-analysis by Cidu et al. (2011), resulting in unique exposure conditions for consumers. To date, studies have focused on the health impacts of bottled water derived from natural fresh water sources with limited to no inclusion of the chemical composition of water from seawater desalination processes and the health effects of long-term consumption of these products. Desalination technologies efficiently remove many trace elements and natural ions from product water potentially reducing the risk of exposure to trace contaminants but also reducing the intake of essential nutrients (Ludwig 2010; WHO 2011a). The inadequate intake of essential minerals from prolonged consumption of desalinated water may result in nutrient deficiencies and profound health impacts for dependent populations (Guler & Alpaslan 2009).

The aims of this multi-element study were to: (1) quantify the trace element and micronutrient content in desalinated and non-desalinated bottled water available to consumers in Qatar; (2) assess the impact of container type on trace element content; and (3) use the National Academy of Sciences Institute of Medicine (IOM) recommended daily allowance (RDA) and adequate intake (AI) values to assess the contribution of desalinated and non-desalinated bottled water consumption to Ca, Mg and F^– intakes for adults and children.

Bottled water samples (n = 56) were obtained from eight local markets in Qatar in April 2013. The samples included both domestically produced and imported brands from 15 countries and constituted a comprehensive representation of the bottled water types available to consumers in Qatar. Desalinated water produced in Qatar, United Arab Emirates (UAE), Kuwait and Bahrain constituted 38% of all samples collected. Qatari brands represent the highest percentage (20%) of samples collected.

The majority of samples (86%) were packaged in polyethylene terephthalate (PET/PETE) containers with only eight samples bottled in glass containers.

Non-acidified sample aliquots were used for all analyses excluding quantification of trace elements. Physical properties (pH, electrical conductivity and total dissolved solids (TDS)) were measured at ambient temperature using a Thermo Scientific Orion 5-Star Plus Benchtop Multiparameter Meter in Qatar Environment and Energy Research Institute (QEERI) laboratories, Doha, Qatar. Total concentrations of anions (Br^–, Cl^–, F^–, NO₃^– and SO₄^2–), major elements (Ca, K, Mg and Na), HCO₃^– and total organic carbon (TOC) were quantified at the Institute of Earth Sciences, Heidelberg University, Heidelberg, Germany. Anions were analyzed using an ion chromatograph (Dionex ICS-1100), major elements were analyzed using inductively coupled plasma optical emission spectrometry (Agilent 700) and HCO₃^– and TOC were analyzed with TOC-V_CPH Analyzer (Shimadzu).

For trace element analysis, samples were acidified with HNO₃ (Normaton for trace analysis) to pH < 2 for quantification of total As, Ag, Ba, Be, Cd, Co, Cr, Cu, Fe, Mn, Mo, Ni, Pb, Sb, Se, Th, Tl, U and Zn using inductively coupled plasma mass spectrometry (Agilent 7500ce and 7700x) at the Laboratory of Bioinorganic Analytical and Environmental Chemistry, University of Pau, France. Analysis was performed in accordance with ISO 17294. One laboratory blank was included in each analysis set of 10 samples and a certified reference material (NIST 1643e) was included in each batch to ensure optimum instrument performance. All elemental results for NIST 1643e were within 10% of the certified values indicating satisfactory method performance.

Trace element content and physical properties, anion and major element content are displayed in Tables 1 and 2, respectively. Concentration ranges for a given element were limited from only 10² (As, Ag, Ba, Be, Cd, Co, Fe, Pb, Sb, Th and Tl) to 10⁵ (Cr, Cu, Mn, Mo, Ni, Se, U and Zn). Mn displayed the greatest range in concentration among all samples (approximately 10⁵) whereas Cd and Be displayed the smallest (approximately 10²). All samples were below the United States Environmental Protection Agency (USEPA), World Health Organization (WHO) and GCC Standardization Organization (GSO) guidelines for drinking water with the exception of Ni (three samples exceeded the GSO limit), Th (one sample exceeded the USEPA limit) and NO₃ (two samples exceeded the USEPA limit). Overall, the results indicate bottled water consumption is not a significant route of exposure for As, Ba, Be, Cd, Cr, Pb, Sb and U for populations in Qatar. Guideline values for Ag, Co, Fe, Tl or Zn are currently unavailable; however, concentrations ranged within expected levels for these elements.

Physical parameters and concentrations of major ions did not notably vary between the non-desalinated water brands, with the exception of select parameters. Samples from Germany contained significantly higher mean concentrations of TDS (p = 0.001) and Be (p = 0.003), while bottled water imported from France contained significantly higher mean As concentrations (p = 0.001). Samples from the United Kingdom contained significantly higher mean concentrations of nitrate (p < 0.001) and those from Scotland contained significantly lower mean pH levels (p = 0.003).

Variations observed between desalinated and non-desalinated bottled water samples were particularly interesting in regards to their potential impacts on consumers. Desalinated water produced in Qatar, UAE, Kuwait and Bahrain had significantly (p < 0.05) lower levels of the essential ions Mg and Ca and the non-essential trace elements As, Co, Ba and U when compared to the non-desalinated samples (Table 3). The mean Ca concentration for the desalinated bottled waters (11.3 mg/L) was significantly less than the WHO minimum requirement of 20 mg/L (p < 0.001) whereas the mean concentration from the non-desalinated waters was 57.5 mg/L, within the optimum range. The same was observed for Mg; desalinated bottled water contained an average of 10.9 mg/L while the mean concentration measured in the other bottled water samples was 18 mg/L. The low major and trace element content observed in desalinated water samples is indicative of the effective removal of dissolved inorganic constituents by the thermal desalination technologies currently used by the Gulf nations. While the removal of non-essential trace elements from drinking water is protective to consumer health, the removal of essential nutrients presents unique health risks.

Drinking water can contribute up to 20% of the daily requirements for Ca and Mg, providing an important dietary supplement, and is a significant source of F^– (WHO 2005; Galan et al. 2002; Avni et al. 2013). As Ca and Mg are present as free ions in water, they may be more readily absorbed and have increased bioavailability in drinking water over other food sources (Galan et al. 2002; Avni et al. 2013). Studies suggest Ca and Mg deficiencies are associated with an increased risk for a number of cancers including kidney, colon and lung (Yang et al. 1997, 1998; Chen et al. 2012). Ca deficiencies have also been associated with rickets in children and osteoporosis in adult women (WHO 2005). In addition, decreased heart health has been linked with dietary Mg deficiencies (Chiuve et al. 2013; Del et al. 2013). Kousa et al. (2006) directly linked Mg deficits in drinking water with increased risk of acute myocardial infarction showing that a 1 mg/L increase in drinking water Mg levels reduced the risk by 4.6%. Fluoride deficiencies have been associated with long-term consumption of desalinated water resulting in increased oral health risks such as dental caries, especially in children less than 9 years of age (WHO 2005; Pizzo et al. 2007).

There is evidence of dietary Ca and Mg deficiencies in many parts of the world today; these deficiencies may become increasingly exacerbated as reliance on desalinated water increases (WHO 2011a). The contribution of drinking water to Ca, Mg and F^– intake for populations dependent on desalinated water is explored in the following section.

Drinking water contributes between 1 and 5% of the total dietary intake of several micronutrients. The contribution is much higher for Ca and Mg, up to 20%, as well as F^–, up to 80% of the daily dietary requirement (WHO 1996, 2005). Table 4 shows the IOM AI_Ca, AI_F and RDA_Mg for adult males and adult females, including those who are pregnant, and the contribution (given as percent) of desalinated and non-desalinated bottled water consumption to the recommended intakes.

Desalinated bottled water contributed a significantly lower percentage of the AI_Ca (p = 0.001) and RDI_Mg (p = 0.016) for adults when compared to non-desalinated water sources. The contributions were not significantly different for F^– at the α = 0.05 level, however, they were notable, with desalinated bottled water contributing approximately 5 times less than the other types. For adult males, non-desalinated bottled water contributed approximately 14% of the AI_Ca and 15% of the RDA_Mg, whereas desalinated brands only accounted for approximately 3% of AI_Ca and 6% of the RDA_Mg. The same was observed for adult females, both pregnant and non-pregnant. The deficit in Ca and Mg content in desalinated seawater increases the risk of Ca and Mg deficiencies among populations dependent on desalinated water. Beyond direct contribution to nutrient intake, cooking with water of low mineral content has been shown to decrease mineral content in cooked foods; therefore, the low Ca and Mg content in desalinated water may indirectly impact the dietary intake of these elements and further increase the risk of deficiencies (Haring & Van Delft 1981).

Children are uniquely susceptible to mineral deficiencies and are at higher risk of developmental impacts (WHO 2005). Table 5 shows the contribution of desalinated and non-desalinated bottled water to the AI_Ca, RDI_Mg and AI_F for children ages 1–10 years. Desalinated bottled water contributed between 2.7 and 3.8% of the AI_Ca while non-desalinated brands ranged from 11.3 to 15.3%. For the RDA of Mg, desalinated water provided between 3.4 and 10.3% while non-desalinated types provided between 8.9 and 26.7%. This discrepancy was also present for the adequate intake of F^–; desalinated water contributed between 3.2 and 9.1% while non-desalinated bottled water contributed up to 53.3% of the AI_F.

Owing to social and cultural preferences for bottled water, children in many countries throughout the Gulf rely on bottled water as their primary drinking water source. The nutrient requirements for adolescents place them at higher risk of health complications from dietary deficiencies. Insufficient intake of Ca during puberty increases the risk of bone fracture and osteoporosis in adulthood as approximately 40% of total bone mass is accumulated during this stage (Greer & Krebs 2006). Low dietary intake of Mg has shown decreased lung function among children, particularly among those with asthma and other respiratory diseases (Gilliland et al. 2002). Given the low mineral content in desalinated drinking water, children relying on this water source may require dietary supplements to minimize health risks.

Samples were bottled in either glass (n = 8) or PET/PETE (n = 48) containers. The majority of samples were still water (n = 47), only nine samples were carbonated. Water stored in glass bottles contained significantly higher concentrations of select essential and non-essential parameters including Be (p = 0.001), Mn (p = 0.011), As (p = 0.038), Ba (p = 0.0006), Tl (p = 0.025) and Pb (p = 0.039) (Table 6).

Studies comparing glass and PET bottle types have varied in their observations. Cidu et al. (2011) did not observe significantly higher trace element concentrations in natural mineral waters bottled in glass containers; however, several studies have found higher levels of trace elements including K, Li, Na, Pb, U and Zr in glass bottles compared to plastic containers (Misund et al. 1999; Shotyk & Krachler 2007b). All but one glass bottled sample contained carbonated drinking water; the impact of each variable (e.g. container type or CO₂) on trace element content requires further investigation.

Many countries in arid or semi-arid regions already rely on alternative water resources (such as treated wastewater and desalinated seawater) to meet growing water demands. The Gulf region is in a state of acute water stress and depends solely on seawater desalination for potable water. The health implications of this dependency have been poorly researched thus far. This study was prompted by Qatar's dependence on seawater desalination and preferences for bottled water by local populations.

Trace element content in bottled water samples from local markets within Qatar did not exceed drinking water guidelines of the USEPA, WHO or GSO. Results indicate that Qatar's population is not at elevated risk of dietary exposure to As, Ba, Be, Cd, Cr, Pb, Sb and U from consumption of both desalinated and non-desalinated bottled water types available in the country. Consumers who primarily consume desalinated water brands further minimize risk of exposure to As, Ba, Co and U as levels were significantly lower than in non-desalinated bottled water.

Desalinated bottled water contained significantly lower levels of Ca, Mg and F^– than other water types. This deficit may increase the likelihood of nutrient deficiencies among populations dependent on desalinated water as results indicate desalinated bottled water consumption was not a significant contributor to recommended daily intakes for Ca, Mg and F^– for adults and children.

Drinking water guidelines should reflect the low mineral quality of finished water and impose minimum nutrient requirements to minimize adverse human health impacts. A variety of cost-effective remineralization techniques are currently available to increase nutrient content of desalinated seawater.

This article was made possible by an NPRP (National Priorities Research Programme) award (5-126-1-030) from the Qatar National Research Fund (a member of The Qatar Foundation). The statements made herein are solely the responsibility of the authors. We would like to acknowledge Islam Qunnaby for his contributions to the project. The authors are appreciative of the laboratory teams at University of Heidelberg (Stefan Rheinberger, Christian Scholz and Silvia Rheinberger) and University of Pau (Dr Hugues Preud'Homme) for their support.