Lead and copper in drinking water fountains-information for physicians.

Background: Lead and copper are potentially toxic metals. The objective of our work was to test the water from the drinking fountains of a large public access office complex in southwest Houston, Texas, for the presence of lead, copper, and microbiologic contamination. The data for the water fountains were compared with what we found in the local municipal drinking water supplies.

Methods: Samples were collected as the first draw at the beginning of the work week. These samples were acidified to prevent the precipitation of heavy metals and analyzed using United States Environmental Protection Agency (USEPA) approved procedures and quality control.

Results: Traces of lead were detected in 37.5% and copper in 100% of the tested water fountains. In two buildings, concentrations in some fountains exceeded the USEPA action level for lead (by up to 12-fold) and for copper (by up to 3.9 fold). One sample was positive for total coliform and Escherichia coli bacteria. Comparison with samples from the local municipal drinking water supplies indicated that both metals and bacteria were the result of secondary contamination at the water fountain sites.

Conclusions: This study showed that drinking water fountains can be an unexpected and unappreciated source of intake of metal and bacterial contaminants.

Key Words: drinking water fountains, total lead, total copper, Escherichia coli, corrosion, contamination, municipal water supplies.

**********

Over the last three decades, lead and copper have received growing attention from the scientific community, regulatory agencies, and the public. (1-6) Lead is a toxic metal that in the past was widely used in household products, paint pigments, plumbing fixtures, as well as a gasoline additive. Even at trace concentrations, lead can cause a variety of undesirable health effects, especially in children six years and younger and in pregnant women. (4,7,8) Lead can also cause fetal abnormalities, as it is mobilized from bone deposits during pregnancy and can cross the placenta. (4,6) Toxic effects can include anemia, mental retardation, neuropathology, and physical organ damage. (3,9) Susceptibility to lead poisoning is related to levels of calcium, iron, phosphorous, vitamins A and D, dietary protein, and alcohol consumption.

Ingestion of lead-based paint chips found in older homes and lead-contaminated soil are well-known sources of lead exposure of inner city children. (10) However, drinking water must not be ignored as a source of lead exposure to children and adults. The United States Environmental Protection Agency (USEPA) estimates that about 20% of human exposure to lead is attributable to drinking water sources. (11)

Copper is a malleable metal that currently is widely used in place of lead in household plumbing systems and fixtures. (6) Although copper is a micronutrient, in excessive concentrations, it too, is toxic. (5,12) Health problems can range from gastrointestinal and respiratory irritation, to liver, kidney, and brain tissue injury. Sparks and Schreurs (13) found that [beta]-amyloid formation was increased in laboratory animals who ingested copper-containing water, but not in animals ingesting copper-free water. Moreover, a link has been proposed between the dietary intake of copper by the elderly human population and increased [beta]-amyloid formation, which has been associated with neurodegenerative conditions, such as Alzheimer disease. (14-16)

In the United States, federal and state governments have made the reduction of metal exposure a priority. Beginning in the 1980s, the regulatory measures included a phase-out of leaded gasoline, reduction of lead in public drinking water by safeguarding surface watersheds and ground water, reduction of lead in the air by controlling industrial pollution, and restrictions on the use of lead in consumer products. The recent USEPA Lead Awareness Program was launched to protect human health and the environment through regulations, research, and public education. (10)

Drinking water, normally, does not contain lead or copper when the water leaves the municipal treatment plants. Concentrations of these metals are regulated under the authority of the Safe Drinking Water Act (SDWA) of 1974 and its amendments. (17) The USEPA sets the maximum contaminant level goal for lead at zero. For copper, the goal is set at concentrations not to exceed 1.3 mg/L (1,300 [micro]g/L), based on currently available technology. However, as water travels within the distribution system, lead and copper may leach into the previously treated water as secondary contamination via contact with corroded lead pipes or copper pipes containing lead solder. Homes built before 1986 are more likely to have lead pipes and lead solder in the plumbing fixtures.

Recognizing this situation, the USEPA issued the Lead Reduction Rule. (18) This rule requires water utilities to control the corrosiveness of water and, as needed, replace lead service lines used to carry water from the street to the home. The USEPA requires that the total lead in drinking water not exceed 15 [micro]g/L and total copper not exceed 1,300 [micro]g/L.

Nonetheless, the safety of drinking water cannot be taken for granted, and incidents of water contamination continue to surface at the consumer's tap. Episodes of heavy tap water contamination by infectious agents associated with refrigerated water fountains were reported in hospitals, nursing homes, dental-unit water, and the day-care environment in studies conducted in France, Spain, USA, and Canada. (19-25) Herein, we report findings of elevated levels of lead and copper in the drinking water fountains of large public access facilities located in the city of Houston, Harris County, Texas.

Methods

On February 21, 2005, 45 drinking water fountains located in five public access buildings in southwest Houston were tested for concentrations of total lead, total copper, total coliform bacteria, and E. coli. The selection of drinking fountains for the survey was based on the known location of fountains in five buildings, provided by the organization's safety officer. For the purpose of this paper, these buildings are identified as Buildings #1, #2, #3, #4, and #5. Nine drinking water fountains (20.5%) were tested in Building #1; 12 (27.3%) in Building #2; 5 (11.4%) in Building #3; 5 (11.4%) in Building #4; and 14 (29.5%) in Building #5. The five buildings in the same complex were selected to represent structures of different ages, from as new as 1 year (Building #4) to more than 50 years (Building #1), with buildings #2, 3, and 5 being of intermediate ages--29, 44, and 30 years, respectively.

Samples were obtained at the same time, at the start of the work week and work day. Investigators used identical 60 mL new polyethylene containers to collect water samples, following an approximately 10 second purge run. The bottles were spiked with 0.6 mL of 1% high purity nitric acid to prevent the heavy metals from precipitating. Samples were labeled according to location and date. Other recorded information included water fountain manufacturer, model, condition, and presence of water filters.

The concentrations of total lead and total copper were determined by atomic adsorption spectrometry. (26) Experienced University of Texas-Houston School of Public Health analysts tested the water samples using the Varian Spectra AA-10/20-Plus Spectrometer fitted with a graphite furnace. Samples were analyzed according to USEPA Method 220.2 for copper and Method 200.9 for lead.

The linear range of absorbance for lead (with ammonium dihydrogen phosphate a matrix modifier) was determined to be between 2 and 20 [micro]g/L, with a minimum detection limit (MDL) of 1 [micro]g/L. For copper, the range was from 5 to 40 [micro]g/L, with an MDL of 4 [micro]g/L. For each metal, a five-point calibration curve was constructed (blank, 2, 5, 10, and 20 [micro]g/L for lead and blank, 5, 10, 20, and 40 [micro]g/L for copper), and samples absorbing within the appropriate ranges were quantified by linear regression. Samples that absorbed over the calibration range were serially diluted and rerun. We performed tests for coliform bacteria and E. coli by the Coliert[R] presence-absence technique, (27) after holding the samples in an incubator for 18 hours at 35[degrees]F.

Results

Of the 45 water fountains intended for sampling, one in Building #2 and four in Building #5 were not operational. The majority (80%) of the in-service water fountains were manufactured by the Halsey-Taylor Co. Four fountains (10%) were Oasis[R], four (10%) were Elkay[R], and two (2.5%) were Aqua Chill[R] brand. Traces of copper were found in the drinking water of all 40 sampled fountains and traces of lead were found in 15 (37.5%).

Concentrations of total lead ranged from nondetectable to as much as 210 [micro]g/L (Table). The median concentration of lead for all samples was below the limit of detection (<1 [micro]g/L), but the mean and the standard deviation were 16 [+ or -] 48 [micro]g/L. Concentrations of total copper ranged from 2 [micro]g/L to as much as 5,043 [micro]g/L. The median concentration was 320 [micro]g/L and the mean 650 [+ or -] 958 [micro]g/L. Compared with the action levels for lead and copper, the water fountains in the west wing of the 3rd and 7th floors of Building #5 had excessive concentrations of both metals, amounting to 180 [micro]g/L of total lead plus 3,226 [micro]g/L of total copper on the 3rd floor and 150 [micro]g/L of total lead plus 5,043 [micro]g/L of total copper on the 7th floor. The water fountain in the east wing of the 6th floor of the same building had an excessive concentration of total copper, 1,836 [micro]g/L. One of the water fountains in the basement of Building #5 tested positive for total coliform bacteria and E. coli. Building #3 had excessive lead concentrations in the three fountains located in the basement and on the 1st floor, being 210, 26, and 36 [micro]g/L for total lead, and in addition, 1,671 [micro]g/L total copper was detected in one of these fountains.

Figures 1 and 2 show the distribution of lead and copper concentrations according to the building age and the presence/absence of water filters. The lead concentration was the greatest in buildings that were 30 to 40 years old and was lower in the buildings that were either newer or older. The concentration of total lead was significantly correlated with the concentration of total copper (r = 0.69, P < 0.03).

Discussion

The Safe Drinking Water Act (SDWA) that was promulgated by the United States Congress in 1974 gave the authority to the USEPA to set and enforce national drinking water standards. (17) The 1986 amendments to the SDWA included the prohibition against the use of lead-based fluxes, solders, and pipes in the public water systems. (28) In 1988, the US Congress passed the Lead Contamination Control Act, (29) which, among other matters, dealt with the recall of lead-lined drinking water coolers. The 1996 Amendments to the SWDA emphasized the public's right to know about drinking water safety. The Act allocated federal funds for upgrading drinking water systems and called for the greater protection of water supplies. (30)

Together, the USEPA, state governments, and water utilities work to ensure that these standards are met, and tremendous success has been achieved nationwide through these regulatory activities. However, one cannot presume that the drinking water is always safe, as a number of threats to water quality continue to exist. Specifically, as treated municipal water moves through the poorly designed and poorly maintained distribution systems, it can become contaminated by lead, copper, and microbiologic agents.

Recognition of these problems is evident from a recent USEPA news release. (1) On March 7, 2005, the USEPA announced the initiation of the Drinking Water Lead Reduction Plan. This action, which followed a USEPA assessment of the current implementation of federal and state lead regulations, was launched to improve monitoring, treatment, service line management, and customer awareness. The USEPA plan specifically singled out the issue of lead in tap water in schools and childcare facilities, thereby calling for better protection of these vulnerable populations.

Our study addressed still another issue involving lead; that is, the contamination of water derived from drinking water fountains. Drinking fountains, which are free standing or wall-mounted chilled drinking water dispensers, have been in use for nearly a century. They were originally developed to protect the public health and prevent the infection spread by sharing a drinking cup. Some of the commonly used designs date back to 1912. More modern designs use lead-free materials, but they use copper in cooling components.

We reasoned that the best time to test for lead and copper would be after the water fountains had not been used, ie, over the weekend. With this consideration in mind, our samples were collected in the morning at the beginning of the work week, specifically on Monday, between 8 and 9 AM.

We found that water in several fountains contained both lead and copper. At 150 and 180 [micro]g/L of total lead, these water fountains exceeded the USEPA action level for this metal by 10- to 12-fold. At 3,226 and 5,043 [micro]g/L, the same water fountains exceeded the action level for copper by 2.5 to 3.9-fold. We also found bacterial contamination in one fountain. The number of contaminated water fountains perhaps might have been even greater if we had access to the five water fountains (4 in Building #5 and one in Building #3) that were removed from service before our sampling.

[FIGURE 1 OMITTED]

The building age was not a linear factor in the concentrations of lead and copper (Fig. 1), but rather, concentrations varied with the age of installation, the type of water fountains, and the material used in the cooling system. This was illustrated by the example of Building #5. The lead and copper concentrations in this building were not found to be high in water fountains above the 7th floor, although concentrations over the EPA action levels were found in the lower floors. While the building proper was constructed in 1975, the upper floors (and associated water fountains) were added more than a decade later.

[FIGURE 2 OMITTED]

The presence of filters seems to be associated with lower concentrations of total lead, but it did not have an apparent effect on concentrations of total copper (Fig. 2). However, further studies would be required to better clarify the role of the filters.

The findings of lead and copper in the water fountains in the five public access buildings in Houston should be viewed in contrast to the data from the local municipal water supplies. In February 2004, the year before the present study of the drinking water fountains, we conducted a sampling of the municipal tap water throughout Houston/Harris County, Texas (previously unpublished field data). The entire Harris County has an area of 1,777.89 square miles. We divided the county into 37 equal sized quadrants, 48 [mi.sup.2] each, or roughly 0.1 degrees radian, using a "Grid" command of the geographic information system, (GIS) computer-assisted mapping software by MapInfo, Inc. (31) Samples of tap water from public access buildings (stores, restaurants, movie houses) were collected from each quadrant on the same day. Water fountains were explicitly excluded from the year 2004 sampling. The sampling and laboratory analytical methods used in this past study were the same as for the water fountain study. Concentrations of total lead in this study were all below the limit of detection for our method (0.1 [micro]g/L) and therefore, not shown. The median concentration of total copper was 37 [micro]g/L, and the mean was 87 [micro]g/L [+ or -] 110 [micro]g/L.

Figure 3 shows the spatial distribution of the total copper in the tap water sampled in Harris County in 2004, using two and three-dimensional surface trend analyses. This map was created with the GIS software "SURFER," version 8, by Golden Software, Inc. (32) Concentrations varied from none detected to 490 [micro]g/L, observed primarily near the northern and southern boundaries of Harris County. These areas corresponded to where recent urban development has encroached into land previously used for agriculture and where copper-based agro-chemicals might have been applied to crops. Another potential (and more likely) explanation is that more copper plumbing is used in the newer structures, built after 1986.

[FIGURE 3 OMITTED]

Although copper was detected in the tap water in some peripheral outlying areas of Harris County, at no location did the concentrations of copper in municipal water samples exceed 1,300 [micro]g/L, the USEPA action level for copper. Therefore, we are reasonably certain that, overall, tap water supplies in Houston/Harris County are treated sufficiently well for lead and copper.

From this earlier study of the municipal drinking water in Houston/Harris County, we expected concentrations of total lead in the water distributed to the fountains to be below the detection limits and the concentration of copper to be between nondetectable and no greater than 50 [micro]g/L. The exceptionally high concentrations of both metals, which we found as a first draw in several drinking water fountains in the public buildings, clearly suggests a secondary, local contamination source, likely due to corrosion in the fixtures and plumbing directly at these fountains.

Our samples were of first draw lead and copper levels. Flushing of water fountains to reduce lead and copper concentrations is recommended, but this is hardly a permanent solution. A study in New Jersey investigated the effectiveness of flushing on reducing lead and copper levels in school drinking water. First-draw lead and copper levels (median concentrations of 10 [micro]g/L and 260 [micro]g/L, respectively) decreased significantly after 10 minutes of flushing, but concentrations had increased significantly by lunch time (median concentrations of 7 [micro]g/L lead and 120 [micro]g/L copper), after normal use of the fountains by students. (33)

Summary

This study clearly shows that drinking water fountains might be an unanticipated, unappreciated, and unnecessary source of lead and copper intake. The findings underscore the necessity for the proper maintenance of water supply equipment, as well as the testing of water quality, to avoid exposure of the public to metal and bacterial contaminants. The fact that elevated levels of lead and copper were found at the water fountains in excess of the action levels did not imply that any laws were violated. It simply meant that corrective actions were required, and indeed were subsequently taken, in the affected buildings to prevent risks to the public. Retrofitting of water fountains with filters of newer models could be attempted for lead control, but further study is needed on the effectiveness of filters. (6)

Acknowledgments

The authors gratefully acknowledge the contribution to this project of several graduate students at the University of Texas School of Public Health, including Ruja Abdo, Vijay Aluri, Sam Annor, Consuelo Ramirez, Ami Shah, Dong Sun, and Yun Xing, who participated in the collection of the water samples.

References

1. United States Environmental Protection Agency: News Release: EPA to Strengthen Protection from Lead in Drinking Water. March 7, 2005.

2. Howson C. Hernandez-Avila M, Rall DE, Committee to Reduce Lead Exposure in the Americas. Lead in the Americas: A Call for Action, 1st ed. Washington, DC, Board on International Health, Institute of Medicine, 1996.

3. Hammond PB, Bornschein R, Zenick H. Toxicological considerations in the assessment of lead exposure. Neurotoxicol 1984;5:53-66.

4. Agency for Toxic Substances and Disease Registry: Toxicological Profile for Lead. Atlanta, GA, Department of Health and Human Services, ATSDR, 1999.

5. Agency for Toxic Substances and Disease Registry: Toxicological Profile for Copper. Atlanta, GA, Department of Health and Human Services, ATSDR, 1999.

6. United States Environmental Protection Agency: Consumer Fact-Sheet on Copper. USEPA Office of Ground Water & Drinking Water, 2005.

7. Canfield R, Henderson C, Cory-Slechta D, et al. Intellectual impairment in children with blood lead concentrations below 10 [micro]g per deciliter. New Engl J Med 2003;348:1517-1529.

8. Bernard S. Should the Center for Disease Control and Prevention's childhood lead poisoning intervention level be lowered? Am J Pub Hlth 2003;93:1253-1260.

9. Meyer P, Pivetz T, Dignam T, et al. Surveillance for elevated blood lead levels among children-United States, 1997-2001, MMWR 2003;52:1-21.

10. United States Environmental Protection Agency: Lead in Paint, Dust, and Soil. EPA, the National Lead Information Center, Washington, DC, 2004.

11. United States Environmental Protection Agency: Lead in Drinking Water: Basic Information. USEPA Office of Ground Water & Drinking Water, 2005.

12. Georgopoulos PG, Roy A, Yonone-Lioy MJ, et al. Environmental copper: its dynamics and human exposure issues. J Toxicol Environ Hlth 2001; Part B: 341-394.

13. Sparks DL, Schreurs BG. Trace amounts of copper in water induce b-amyloid plaques and learning deficits in a rabbit model of Alzheimer's disease. Proc Natl Acad Sci 2003;100:1-5.

14. Jansson ET. Alzheimer disease is substantially preventable in the United States-review of risk factors, therapy, and the prospects for an expert software system. Med. Hypoth 2005;64:960-967.

15. Marx J. Possible role for environmental copper in Alzheimer's disease. Science 2003;301:905.

16. Williams JH, O'Connell TC. Differential relations between cognition and (15) N isotopic content of hair in elderly people with dementia and controls. J Gerontol Medl Sci 2002;57:797-802M.

17. United States Environmental Protection Agency: Drinking Water Standards. USEPA, Office of Ground Water & Drinking Water, 2005.

18. United States Environmental Protection Agency. Lead in drinking water: lead and copper rule. Federal Register, 1991, 56 RF 26460.

19. Lalande V, Barbut F, Varnerot A, et al. Pseudo-outbreak of Mycobacterium gordonae associated with water from refrigerated fountains. J. Hosp Inf 2001;48:76-79.

20. Usera MA, Echeita A, Aladuena A, et al. Investigation of an outbreak of water-borne typhoid fever in Catalonia in 1994. Enferm Infecc Microbiol Clin 1995;13:450-454.

21. Atlas RM, Williams JF, Huntington MK. Legionella contamination of dental-unit waters. Appl Environ Microbiol 1995;61:1208-1213.

22. Morera MA, Espejo E, Coll P, et al. Epidemic outbreak of shigellosis following water intake. Enferm Infecc Microbiol Clin 1995;13:160-165.

23. Butz AM, Fosarelli P, Dick J, et al. Prevalence of rotavirus on high-risk fomites in day-care facilities. Pediatrics 1993;92:202-205.

24. Le Querler, L, Donnio PY, Poisson M, et al. Isolation of Escherichia vulneris in drinking water. Ann Biol Clin 1997;55:33-35.

25. Dutka BJ, Walsh K, Ewan P, et al. Incidence of Legionella organisms in selected Ontario (Canada) cities. Sci Total Environ 1984;39:237-249.

26. United States Environmental Protection Agency: Statement of Work for Inorganic Analysis. USEPA, ILM04, 1995.

27. 40 CFR, Part 141: National Primary Drinking Water Regulations. Analytical Techniques, Coliform Bacteria; Final Rule. Federal Register, 1992;57:24744.

28. Environmental Protection Agency: Safe Drinking Water Act Amendments. Washington, D.C.: USEPA Office of Ground Water and Drinking Water, 1986.

29. Centers for Diseases Control and Prevention: Implementation of the Lead Contamination Control Act of 1988. MMWR 1992;41:288-290.

30. Environmental Protection Agency: Safe Drinking Water Act Amendments. Washington, DC, USEPA Office of Ground Water and Drinking Water, 1996.

31. MapInfo, Inc: MapInfo, version 7.5, Troy, NY, 2001.

32. Golden Software Inc: Surfer, version 8, Golden, CO, 2002.

33. Murphy EA. Effectiveness of flushing on reducing lead and copper levels in school drinking water. Environ Hlth Perspect 1993;101:240-241.</p> <pre> All sanity depends on this: that it should be a delight to feel heat strike the skin, a delight to stand upright, knowing the bones are moving easily under the flesh. --Doris Lessing </pre> <p>Irina Cech, PHD, Michael H. Smolensky, PHD, Masoud Afshar, MS, Gregory Broyles, BS, Matthew Barczyk, BS, Keith Burau, PHD, and Robert Emery, DRPH

From the University of Texas, Houston, School of Public Health, Houston, TX.

Reprint requests to Dr. Irina Cech, the University of Texas School of Public Health, 1200 Herman Pressler, Houston, TX 77030. Email: [e-mail]

This study was conducted with the approval of the Committee for the Protection of Human Subjects, University of Texas Houston Health Science Center, and with the approval of the Office of Environmental Health and Safety, the University of Texas Houston Health Science Center.

Accepted November 2, 2005.

RELATED ARTICLE: Key Points

* Water fountains, which are the freestanding or wall-mounted chilled drinking water dispensers, have been in use for almost a century.

* Water fountains were originally developed to protect the public health and prevent people from spreading infection through sharing a drinking cup.

* Lead, copper, and bacteria were found to be present in several drinking water fountains at or above the levels expected in public water supplies.

* Comparison with samples from municipal drinking water in the same area indicated that both metals and bacteria were acquired as a secondary contamination directly at the water fountains.

* The findings underscore the necessity for proper maintenance and testing of water fountains, with corrective actions to prevent risk to the public.

* Physicians need to be aware that drinking water fountains can be an unexpected and unnecessary source of lead, copper, and bacteria in public access buildings.

Table. Concentration of total lead and total copper ([micro]g/L) in
drinking water fountains in five public access buildings, Houston, TX
(February 2005)

       Valid                    Median         Mean
       No. of        % samples  (95% CI)       (95% CI)
Metal  observations  detected   [micro]g/L     [micro]g/L

Pb     40             37.5%     < LOD           16
Cu     40            100.0%     320 (175-591)  650 (344-956)

       Min.           Max.           EPA action    No. (%) of
       concentration  concentration  level         samples above
Metal  [micro]g/L     [micro]g/L     ([micro]g/L)  action level

Pb     < LOD            210             15         5 (12.5%)
Cu     2              5,043          1,300         4 (10.0%)

< LOD, Less than limit of detection by laboratory method; EPA,
Environmental Protection Agency.
Pb, lead; Cu, copper.