A cluster of cases of abdominal pain possibly associated with high copper levels in a private water supply.

ABSTRACT

This paper reports on an outbreak of abdominal pain and vomiting in 12 people who worked on a small industrial estate in rural Cheshire (in the United Kingdom). Investigations at the time were unable to identify the cause, although it was noted that the private water supply was poorly managed. Some weeks later, blue crystals were noted precipitating out of a routine water sample. Analysis of the sample gave a copper concentration of 45,110 [micro]g/L. The symptoms complained of in the outbreak cases were compatible with copper poisoning. Further investigations showed that episodes of blue discoloration of the &inking water had occurred in the past. Furthermore, copper pipes in the distribution system showed evidence of internal corrosion.

Introduction

Outbreaks of illness linked to chemical poisoning resulting from consumption of drinking water are relatively uncommon in the United Kingdom (Hunter, 1997). Those that are identified tend to be associated with relatively large public supplies such as the episode associated with aluminium contamination of the water supply at Camelford in 1988 (Rowland, Grainger, Stanwell-Smith, Hicks, & Hughes, 1990), phenol contamination of the River Dee affecting Chester and North Wales in 1984 (Jarvis, Straube, Williams, & Bartlett, 1985), and the river contamination incident near Worcester (Fowle, Constantine, Fone, & McCloskey 1996). By contrast, illness associated with chemical contamination of small and private supplies has not been reported in recent decades. It is not clear whether the absence of reported episodes associated with private supplies merely represents the inability of current surveillance systems to detect such problems. This paper reports an incident of illness most likely associated with elevated levels of copper in a private water supply.

The Outbreak

The outbreak investigation was initiated in November 2000 following complaints made by workers at a factory situated on a self-contained rural estate in Cheshire. A joint health authority/local-authority outbreak team was formed. Epidemiological investigations consisted of interviewing all 10 workers in the affected factory and contacting other units and homes on the site to identify other possible cases on site. The investigations revealed that two more people, working at different units, also had been symptomatic at the same time. All 12 affected people complained of abdominal pain, two complained of vomiting, and two complained of a metallic taste after drinking water. The affected workers normally consumed bottled water, but on a day when they had run out of bottled water, they had drunk tap water. Some had boiled the water, and some had drunk straight from the tap. The complaints were made five days after this event, when symptoms had ceased.

Approximately 350 people work or live on the estate, in a nursing home, a hostel, several industrial units, and some 30 private domestic residences. Each industrial and residential unit was visited by an environmental health officer who asked about symptoms among workers or residents, in the nursing home, the senior nurse was interviewed and then reviewed residents' nursing records.

Further questioning of people working on the estate revealed that many people had stopped drinking the water because they did not like the taste. Several individuals regularly drank tap water, and none of them complained of symptoms. Similarly, no one who did not drink tap water was affected.

The water supply to the estate came from two boreholes. The water was chlorinated by salt electrolysis before distribution throughout the estate. Upon examination of the site, it was concluded that the water supply system was generally old and not well maintained. The seals of the boreholes were poor, and there were cracks on the ground, although it was not established that these cracks permit the ingress of substances into the boreholes themselves (see photo at right). There was an aboveground diesel storage tank in the yard, which has since been removed. A nearby timber treatment plant used a number of chemicals to preserve wood. The boreholes also were in close proximity to farmland.

A number of water samples were taken at the time for biological and chemical tests such as tests for pH, nitrate, nitrite, iron, manganese and copper, zinc, chloride, sulphate, calcium, magnesium, sodium, potassium, silver, cadmium, nickel, silver, chromium, arsenic, mercury, selenium, lead, cyanide, chlorine, and polycyclic aromatic hydrocarbons (PAHs). The only notable abnormality was a marginally elevated nitrate level (mean: 54.2 milligrams per liter [54.2 mg/L] N[O.sub.3]; range: 52.4-56.3) (under European Directives, the current maximum acceptable limit for N[O.sub.3] is 50 mg/L). Microbiological tests were negative, as were tests for organic contaminants. Copper levels at that time were within normal levels, although there was some marked sample-to-sample variation. Of 10 samples analyzed for copper, five had concentrations of <200 micrograms per liter (200 [micro]g/L), and the remaining samples had 300, 309, 319, 559, and 2,550 [micro]g/L. The maximum prescribed concentration of copper in water at the time of the incident was 3,000 [micro]g/L (Private Water Supplies Regulations, 1991). Although not yet implemented for private water supplies, the new European standard for copper is now 2,000 [micro]g/L (European Union, 1998).

At the time of the initial investigation, the outbreak team was unable to reach a conclusion about the cause of the symptoms on the site. The team advised, however, that young babies not drink the water and that the owners of the supply implement repairs to the borehole and water distribution system. Residents and others were kept informed of the investigation on a regular basis.

Some 10 weeks after the incident, during further sampling from the site, it was noticed that blue crystals were precipitating out of a water sample. Analysis of this sample gave a copper concentration of 45,110 [micro]g/L. On the basis of this result, the authors advised the owners of the supply that they were to cease using the borehole and start taking their water from the mains. The change to supply by a local water utility was implemented within 24 hours. When the connection to the public water supply was made, the distribution system was flushed, and blue water was noted coming from some of the taps. Further investigations suggested that this phenomenon had occurred on a number of occasions.

Discussion

There is now increasing evidence that high copper concentrations (about 3,000 [micro]g/L or greater) can cause gastrointestinal upset. This was shown in a randomized control trial in which volunteers received tap water with no added copper; water with 1, 3, or 5 mg per cubic liter of added copper sulphate for a two-week study period; then one week of standard tap water (Pizarro, Olivares, Gidi, Araya, 1999; Pizarro, Olivares, Uauy, Contreras, Rebelo, & Gidi, 1999). A diary was kept of water consumption and gastrointestinal symptoms. Nausea, abdominal pain, and vomiting--but not diarrhea--were significantly related to copper concentrations in water (3 mg/L or greater), it would appear that children have higher susceptibility to copper toxicity (Sidhu, Nash, & McBride, 1995).

Copper in drinking water has also bee associated with more long-term adverse effects, most notably liver cirrhosis in children (Muller et ah, 1999; Price et al., 1996; Scheinberg, & Sternlieb, 1996; Trollmann, Neureiter, Lang, Dorr, & Behrens, 1999). It would appear, however, that elevated dietary intake of copper alone in these children is insufficient to cause of liver damage and that there also needs to be a genetic predisposition (Muller et al., 1999; Scheinberg, & Sternlieb, 1996). The literature also has reports of people complaining of green hair when copper levels in domestic water supplies are high (Lubach, &Becker, 1982; Nordlund, Hartley, & Fister, 1977).

The symptoms complained of in the outbreak reported here are consistent with the symptoms associated with acute copper poisoning. Although the scheme for categorizing levels of evidence associating water with outbreaks of disease was designed primarily for infectious etiologies, the authors believe it can be applied to chemical outbreaks (Tillett, de Louvois, & Wall, 1998). Using this scheme, we would suggest that the outbreak is "possibly" associated with drinking water since the descriptive epidemiology is compatible with a waterborne route and the copper was found in sufficiently high concentrations in the water to cause the symptoms described. It was not possible, however, to definitively diagnose the illnesses.

The outbreak, then, is possibly associated with high copper levels in tap water. Two previous outbreaks of copper poisoning in the United Kingdom occurred in 1956 and 1960 and were due to water being boiled in copper kettles (Hunter, 1997). In contrast, eight outbreaks of copper poisoning affecting 155 people have been associated with tap water in the United States during the years 1992 to 1998 (Barwick, Levy, Craun, Beach, & Calderon, 2000; Kramer, Herwaldt, Craun, Calderon, & Juranek, 1996; Levy, Bens, Craun, Calderon, & Herwaldt, 1998). A number of factors have caused these outbreaks, including accidental addition of acid to the supply, stagnation of supply, new plumbing systems, and inadequate connections. The copper concentrations in affected supplies ranged from as little as 3,300 to over 100,000 [micro]g/L. Indeed the U.S. Environmental Protection Agency now recommends a limit of 1,300 [micro]g/L in drinking water.

An example of such incident is an outbreak that affected a family in northwestern Vermont (Spitalny, Brondum, Vogt, Sargent, & Kappel, 1984). The father and two daughters reported recurrent episodes of vomiting and abdominal pain 10 to 20 minutes after drinking water that was noted to have a blue tint. The water came from a spring and was distributed through copper pipes. The water was acid (pH 5.8) and corrosive (Langelier Index -3.59). The concentration of copper in the water supply was 7,800 [micro]g/L. The bouts of vomiting stopped when the family stopped using tap water for drinking.

The authors were unable to identify why the water had become contaminated with such high concentrations of copper. Samples taken close to the borehole have always had <200 [micro]g/L of copper, suggesting that the source of contamination was within the distribution system. The most likely explanation was that copper from the pipe works was being dissolved during distribution. There were particularly long runs of copper pipe to certain parts of the estate. There was also evidence that high copper concentrations were an intermittent phenomenon. A plausible cause of this would be occasional acidification of the water, which increased its ability to dissolve the copper from the pipe. The authors were unable to identify why the water was acidified, but they suggest that the acidification may be a consequence of poor management practices. It was noted that at the time of the incident the supply was in a poor state of repair and that no written records detailed how the private supply should be managed and monitored.

Whatever the cause of the contamination was, the change from borehole to public supply seems to have resolved the problem. Subsequent water samples did not show elevated copper levels, and other parameters, including pit, complied with U.K. and European legal requirements. Also the residents reported that their confidence in the water quality had improved.

Another aspect of this outbreak was that the people who complained of illness did not normally drink the tap water, preferring bottled water instead. On the day in question, however, everyone had to drink tap water because the bottled water had run out. People who normally drank the tap water did not complain of illness. This observation suggests that people who are regularly exposed to high concentrations of copper in their supply may become tolerant.

It is likely that such incidents are substantially under-reported The symptoms are generally nonspecific, and few doctors are likely to be able to make a correct diagnosis. The numbers of people affected also are likely to be small, so cases are unlikely to be referred to public health officials. Finally, surveillance for chemical-related illness in the United Kingdom substantially lags behind surveillance for microbial-related illness.

This study highlights the fact that private water supplies are generally less well managed than public water supplies. It is perhaps not surprising that the outbreak affected a private water supply. As already mentioned, the supply was found to be in a poor state of repair, and there were no written operating procedures for maintaining the water safety and quality. The incident illustrates the difficulties of adequately maintaining private supplies even where these supplies provide water to potentially vulnerable populations.

There should be increased monitoring of and research into the quality of private and other small rural water supplies and risks they may pose to health, especially in the case of supplies that serve potentially vulnerable populations. The authors suggest, for example, that a prospective study of the health of people consuming water from small private water supplies, similar to the 1991 study Payment and co-authors did for public supplies, is long overdue. There is also a need for descriptive studies of management practices for such supplies and more formal risk assessments. The application of a formal risk assessment would then be able to guide environmental health professionals when they are inspecting such supplies.

From the experience of this incident, the authors suggest that environmental health professionals consider the following points when inspecting private water supplies:

* the number of people supplied and whether there are people who could be classed as particularly vulnerable,

* the general state of repair of any plant used for water treatment and distribution,

* the presence of documents describing the supply with some form of written risk assessment,

* the presence of documents detailing how the supply is to be managed and monitored,

* clear records showing training of individuals responsible for managing the supply, and

* clear records of appropriate routine monitoring (parameters for monitoring would depend on the risk assessment as well as on legal minimum requirements).

Corresponding Author: Professor Paul R. Hunter, School of Medicine, Health Policy and Practice, University of East Anglia, Norwich, NR4 7TJ, United Kingdom. E-mail: [e-mail].

REFERENCES

Barwick, R.S., Levy, D.A., Craun, G.E, Beach, MJ., & Calderon, R.L. (2000). Surveillance for waterborne-disease outbreaks--United States, 1997-1998. Morbidity and Mortality Weekly Report, 49(SS-4), 145.

European Union. (1998). Council directive 98/83/EC on the quality of water intended for human consumption. Official Journal of the European Community, 1-330(5.12.98), 32-54.

Fowle, S.E., Constantine, C.E., Fone, D., & McCloskey, B. (1996). An epidemiological study after a water contamination incident near Worcester, England in April 1994. Journal of Epidemiology and Community Health, 50(1), 18-23.

Hunter, P.R. (1997). Waterborne disease; epidemiology and ecology. Colchester, England: Wiley and Sons.

Jarvis, S.N., Straube, R.C, Williams, A.L.J., & Bartlett, C.L.R. (1985). Illness associated with contamination of drinking water supplies with phenol. British Medical Journal, 290(6484), 1800-1802.

Kramer, M.H., Herwaldt, B.L., Craun, G.E, Calderon, R.L., Juranek, D.D. (1996). Surveillance for waterborne-disease outbreaks--United States, 1993-1994. Morbidity and Mortality Weekly Report, 45(55-1), 1-33.

Levy, D.A., Bens, MS., Craun, G.E, Calderon, R.L., & Herwaldt, B.L. (1998). Surveillance for waterborne-disease outbreaks--United States, 1995-1996. Morbidity and Mortality Weekly Report, 47(SS-5), 1-33.

Lubach, D., & Becker, J. (1982). Green hair. Hautarzt, 33(1), 32-33.

Muller, T., Schafer, H., Rodeck, B., Haupt, G., Koch, H., Bosse, H., Welling, 12, Lange. H., Krech, R., Feist, D., Muhlendahl, K.E., Bramswig, J., Feichtinger, H., & Muller, W. (1999). Familial clustering of infantile cirrhosis in Northern Germany: A clue to the etiology of idiopathic copper toxicosis. Journal of Pediatrics; 135(2 pt. 1), 189-196.

Nordlund, J.J., Hartley, C., & Fister, J. (1977). On the cause of green hair. Archives of Dermatology, 113(12), 1700.

Payment, P, Richardson, L., Siemiatycki, J., Dewar, R., Edwardes, M., & Franco, E. (1991). A randomized trial to evaluate the risk of gastrointestinal disease due to consumption of drinking water meeting current microbiological standards. American Journal of Public Health, 81 (6), 703-708.

Pizarro, E, Olivares, M., Gidi, V., & Araya, M. (1999). The gastrointestinal tract and acute effects of copper in drinking water and beverages. Reviews in Environmental Health, 14(4), 231-238.

Pizarro, E, Olivares, M., Uauy, R., Contreras, E, Rebelo, A., & Gidi, V. (1999). Acute gastrointestinal effects of graded levels of copper in drinking water. Environmental Health Perspective, 107(2), 117-121.

Price, L.A., Walker, N.I., Clague, A.E., Pullen, I.D., Smits, S.J., Ong, T.H., & Patrick, M. (1996). Chronic copper toxicosis presenting as liver failure in an Australian child. Pathology, 28(4), 316-320.

The Private Water Supplies Regulations [Great Britain], Statutory Instrument 1991, No. 2790. London: Her Majesty's Stationary Office.

Rowland, A., Grainger, R., Stanwell-Smith, R., Hicks, N., & Hughes, A. (1990). Water contamination in North Cornwall: A retrospective cohort study into the acute and short-term effects of the aluminium sulphate incident in July 1988. Journal of the Royal Society of Health, 110(5), 166-172.

Scheinberg, I.H., & Sternlieb, I. (1996). Wilson disease and idiopathic copper toxicosis. American Journal of Clinical Nutrition, 63(5), 8428-8455.

Sidhu, K.S., Nash, D.F., & McBride, D.E. (1995). Need to revise the national drinking water regulation for copper. Regulatory Toxicological Pharmacology, 22(1), 95-100.

Spitalny, K.C., Brondum, J., Vogt, R.L., Sargent, H.E., & Kappel, S. (1984). Drinking-water-induced copper intoxication in a Vermont family. Pediatrics, 74(6), 1103-1106.

Tillett, H.E., de Louvois, J., & Wall, P.G. (1998). Surveillance of outbreaks of waterborne disease: Categorizing levels of evidence. Epidemiology and Infection, 120(1), 37-42.

Trollmann, R, Neureiter, D., Lang, 11, Dorr, H.G., & Behrens, R. (1999). Late manifestation of Indian childhood cirrhosis in a 3-year-old German girl. European Journal of Pediatrics, 158(5), 375-378.