Printer Friendly

Assessing health risk in drinking water from naturally occurring microbes.

In April, 1994, in this journal, a comprehensive review of California bottled water by Allen and Darby was published (1). That paper compared and contrasted bottled water with tap water. This review was written to expand and clarify one critical aspect of that paper: the author's thesis that bottled water deteriorates microbiologically during storage and that microbiological changes occurring in the bottle can have adverse health effects. It is necessary to closely examine the nature of bacteria present in bottled water and the dynamics they experience after bottling in order to make appropriate health risk assessment and generate enforceable regulations which will best protect public health.

Allen and Darby (1) presented four basic components to support their contention that water quality is decreased during storage: (i) heterotrophic plate count bacteria (HPC) increase markedly in numbers; (ii) the HPC concentration in the bottle exceeds 500/mL, which represents a health risk; (iii) the immunocompromised are endangered by HPC; and (iv) the lack of residual disinfectant in bottled water compared to tap water contributes to its deterioration.

Heterotrophic plate count bacteria do indeed increase in numbers in a sealed vessel after bottling (2). No drinking water is sterile (3, 4). Both tap water and bottled water have some HPC content regardless of treatment method or residual disinfection concentrations. No treatment process used in mass production of drinking water yields a sterile product - it produces a safe product. Even tap water with high residual chlorine levels have HPC present (5, 6). Once water enters the bottle the small number of HPC present do what they do in nature - they utilize available nutrients and multiply (7, 8, 9,10,11). Because their natural habitat is water, they can utilize the environmental carbon and nitrogen to multiply (8,12, 13). They do so at an extremely low ionic strength and low water temperature. When one species of HPC completely utilizes its particular environmental nitrogen and carbon sources, it dies and is replaced by another species (13). The dead HPC deteriorate returning available carbon and nitrogen, in a different form, to the water. Thus, individual species increase and decrease in numbers over a long period of time. Accordingly, the microbiological "landscape" of bottled water varies widely both in numbers of HPC/mL and species\present over time. Accordingly, sampling a particular bottle will only produce a snapshot of the HPC content at that one point in time. Eventually, all the nutrients are utilized and the HPC counts become very low, or even undetectable. For example, water stored for the army more than thirty years ago has been found to be sterile (14). It is in part because of these natural cycles that it is not rational to write regulations based on any particular concentration of bacteria in the bottle at any specific point in time after the bottle is filled, nor does an expiration date have meaning.

The conditions under which the environmental HPC multiply are much different from the physical/chemical conditions found in humans. Humans are considerably warmer, saltier, and have much higher concentrations of sugar and protein (2,15). The great majority of naturally occurring HPC that multiply in bottled water will not do so under ambient conditions within the human body (16).

Allen and Darby stated that "changes in the bacteriological quality indicated by heterotrophic plate counts were strongly dependent on the initial water quality...." It should be noted, however, that the concentrations of HPC reached in the bottle, and the cycles of multiplication and decrease, are independent of the initial concentrations of HPC entering the bottle. The small number of HPC initially present at the time of bottling will eventually reach concentrations above [10.sup.4]/mL (10). Even though the specific species might differ, an analogous situation occurs in tap water. If tap water is allowed to stand (e.g., in house plumbing overnight, in low-flow dead ends) the chlorine dissipates and the HPC counts increase.

The second component mentioned in the review is that an HPC concentration of 500/mL or above represents a health threat. However, there has never been an association of any particular HPC concentration with health risk (5). Payment, et al., studying reverse osmosis of filtered and non-filtered tap water published some data that showed an association of HPC concentrations with gastroenteritis and other data that did not show an association (17, 18). Calderon, in two EPA-sponsored studies, found no association of HPC concentrations from point-of-use or point-of-entry devices and gastroenteritis (19, 20). In all federal regulations governing drinking water the only mention of any HPC concentration relates to the analysis of coliforms or chlorine, and not in regard to health risk. In the Total Coliform Rule promulgated in 1991, the EPA stated that if a tap water sample had 500 or [TABULAR DATA FOR TABLE 1 OMITTED] more HPC/mL then there could be a suppression of the recovery of a total coliform by lactose-based methods (multiple-tube fermentation and membrane filtration). The Total Coliform Rule also indicates that the Colilert test (also known as the MMO-MUG) would be useful in these cases. This test is refractory to HPC concentrations (21, 22, 23). The FDA, in the preamble to its new proposed bottled water microbiology regulation, also mentions that HPC levels found in bottled water could cause coliform suppression and recommends the Colilert test (24). The second mention of HPC in the regulations relates to chlorine concentrations in tap water. The EPA states that if HPC concentration in tap water is above 500/mL, this may be indicative of decreased free chlorine concentration (25). This situation is not applicable to bottled water since there is no residual disinfectant.

The third component concerning HPC in the review is that "certain pathogenic bacteria may flourish, endangering immunocompromised individuals." An extensive literature exists concerning the flora of bottled water. It has been uniformly found that those bacteria that can multiply in bottled water have evolved to live a lifestyle specifically suited to that kind of aqueous environment, and are not species known to be human pathogens (10, 26, 27, 28). Additional literature exists which indicates that those HPC in bottled water which do multiply in the bottle do not possess appreciable virulence factors associated with human disease (16, 29, 30, 31, 32).

In order to develop an infection, a number of interrelated factors must be in place. The mere presence of an environmental microbe does not mean that it poses an infectious threat even though it may appear in a list of potential pathogens in a textbook. The correct conditions must exist or an infection will not occur. A very particular series of circumstances is necessary in order for a microbe to cause disease.

Infection is defined as tissue damage to the host resultant from microbial action. The surfaces and mucosal membranes of the human body are densely coated with microbes (i.e., the body's normal flora). As Table 1 demonstrates, the density of our normal microbial flora varies from several hundred bacteria per square centimeter to several billion (33, 34). These microbes are not only natural, but serve a protective function. Their very presence precludes attachment and growth of harmful microbes. All life is associated with a normal microbial component. It would be impossible for living systems to lack a natural microbial flora. Natural, untreated water always has a microbial component. Likewise, vegetables, milk, etc., also have a normal microbial component. Each of these "hosts" creates a particular ecological niche for a particular group of natural microbes.

Before examining the nature of the microbes found in water, it is first necessary to differentiate pathogenic from indicator bacteria [TABULAR DATA FOR TABLE 3 OMITTED] and to understand the nature of indicator bacteria. More than 100 years ago it was found that the actual pathogenic bacteria (e.g., Salmonella or Shigella) were present both infrequently and in concentrations too small to be detected in water. Therefore, public health authorities turned to the analysis of indicators, rather than the pathogens themselves, as a measure to determine water potability. The indicators were not necessarily predictors of particular pathogens, but rather general of contamination of the water supply. It was established in the 1890's that Escherichia coli was always found in high concentrations in the colon of mammals. Unlike other related bacteria (the total coliform group), E. coli was not an inhabitant of the environment at large. E. coli was always found in higher concentrations than the pathogens, and also survived in water as long as, or longer than, the pathogens. Therefore, the presence of E. coli in water did not indicate the presence of a particular pathogen, but meant that the water was contaminated by feces. Because at that time it was difficult to test water directly for E. coli, in the early 1900's a related group was substituted. These were bacteria that were similar to E. coli, which were designated "the total coliform group." Advantage was taken of the observation that E. coli and related bacteria fermented the sugar lactose. Therefore, the coliforms, also known as total coliforms, were created as a group of bacteria which constituted a functional definition based on a biochemical characteristic. The term coliform was one of function and not one of taxonomy. Genera which comprise the coliforms include: Klebsiella, Enterobacter, Citrobacter, Serratia, E. coli, and other species of Escherichia. Unlike E. coli, the other coliforms, including non-coli Escherichia, were not found routinely in the colon and were generally found naturally in source water. It was later found that, unlike E. coli, many other coliforms could live and multiply from the breakdown products of vegetation. Therefore, although total coliforms were easier than E. coli to enumerate, they were a considerably less specific indicator of fecal contamination (26, 34). In 1991, the United States Environmental Protection Agency (EPA) promulgated a new Total Coliform Rule as part of the Safe Drinking Water Act. Because of recent advances in analytical technology brought about by the development of the Colilert test (IDEXX Inc., Westbrook, ME), the EPA re-inserted E. coli as a bacterial analyte for the analysis of drinking water. The Colilert test can simultaneously enumerate total coliforms and E. coli directly from either source or drinking water (21, 22, 23). The EPA required that drinking water be analyzed for total coliforms as an overall measure of treatment efficacy. If total coliforms were found, that positive sample had to be tested for E. coli in addition. Moreover, a stringent re-sampling of the site had to occur within 24 hours. The Total Coliform rule defined coliforms as an overall measure of treatment efficiency and E. coli as a measure of public health risk.

ID50 Values of Some Common Gastrointestinal Microbes.

Species of Microbe ID50

Salmonella 10,000 - 100,000
 (decreases with increase in stomach pH)

Shigella 100

Yersinia 1,000 - 10,000

Campylobacter 1,000 - 10,000

Giardia 1 - 10

Hepatitis A virus 100 - 1,000

Heterotrophic plate count bacteria (HPC) have been established as normal inhabitants of water since water first began to be bacteriologically analyzed (4). In 1977, the FDA examined the presence of HPC in bottled water and concluded

Treated water that is sealed in a bottle and which does not contain a residual concentration of a chemical disinfectant can be expected to contain a few bacteria that are natural to the water, and those few bacteria can be expected to multiply in the bottle during storage. This is a natural phenomenon that may occur in water that is processed and bottled under the best of sanitary and processing conditions. This multiplication is usually of a cyclic nature in that, during storage in a sealed bottle, the bacteria at first multiply and then die off. The few surviving cells may once again give rise to increased numbers that in turn die off. This cycle may occur several times. No microbial spoilage or contamination occurs (35).

HPC are defined by the media and culture conditions by which the water sample is analyzed. There is no universal definition of HPC. There have been a myriad of media used and a large variety of incubation conditions, including differences in the temperatures and times of incubation. Each of these conditions markedly changes not only the numbers of bacteria enumerated, but more importantly, the types of bacteria isolated from water samples (5). Until 1985 the most common measure of heterotrophic bacteria was the Standard Plate Count test (SPC). Now, R2A agar is used most commonly to produce an HPC count. R2A agar contains fewer nutrients and a lower ionic strength than SPC agar and was felt to be more analogous to conditions found in the environment (3).

While a good understanding of the natural history and biology of coliforms and E. coli as indicator bacteria has long been developed, it is only recently that the species comprising the natural flora of water, the HPC, have been studied in detail. Therefore, in order to determine if there is a health threat from the natural microbes found in water, it is necessary to understand their lifestyle and virulence characteristics.

The possible threat to human health by infection is governed by the following Infection Relationship, first expressed by the pioneer microbiologist Theobold Smith almost 100 years ago:

Infection = [Number of Microbe(s)] x [Virulence Characteristics]/Immune Status of the Host

Understanding this relationship and its components is critical to the assessment of risk posed by environmental microbes (33, 34). Accordingly, the individual components of the formula as it applies to the HPC will be discussed.

Number of Microbes:

Normally, the surfaces and mucosal membranes of the body are coated with bacteria and fungi, collectively known as normal flora. Each body site has its own type of normal flora. Infection is associated with at least 1,000,000 bacteria per gram of tissue. In order for microbes to multiply to such large numbers, they must have major virulence factors (to be discussed in the next section) and literally assume control over a particular ecological niche of the host by eliminating the normal flora of the site and overcoming the body's defense mechanisms. This microbial ascendancy does not occur when just a single microbe contacts the body. There is a minimum number of microbes with particular virulence factors required (34, 36).

The major components of immune status.

Characteristic Function

Intact skin Anti-microbial; physical barrier

Normal flora Prevents attachment of invaders;
 makes anti-microbial substances;
 detoxifies ingested chemicals

Polymorphonuclear leukocyte Phagocytizes invading microbes and
 kills them

Lymphocytes Responsible for immune surveillance;
 controls viruses, fungi, and many

Antibody Eliminates bacteria with capsules

Stomach Acid destroys most ingested microbes

Important Differences Affecting Microbial Growth Between
Environmental and Human Niches.

Condition Environmental Water Human Water

Temperature 4 to 20 [degrees] C (average) 37 [degrees] C
Sugar None 100 mg/100 mL
Protein None or minimal 6 g/100 mL
Ionic strength Very low 0.9 Molar (0.85%)

There are two definitions associated with the minimum number of microbes required to produce disease. The first is known as Infectious Dose 50, or ID50. This term is the minimum number of microbes required to produce disease in 50% of contacted hosts. The second term is known as the Lethal Dose 50, or LD50. This term represents the minimum number of microbes required to cause death in 50% of contacted hosts. For each species of microbe there is an ID50 and LD50 for each anatomic site (Table 2) (33). Within a given species there may be differences in the ID50 and LD50 values because some microbial soldiers may be more dangerous than others (36, 37).

Each ID50 and LD50 value is for a particular anatomical site. For example, the ID50 for Group A Streptococcus is in the range of several hundred for the posterior oropharynx but is so high for the gastrointestinal tract that one would virtually have to eat a slurry of bacteria in order to infect the alimentary canal. Likewise, Pseudomonas aeruginosa has an ID50 in the order of 1,000 for burned skin but its ID50 for intact skin is too high to pose a threat. Similarly, the P. aeruginosa ID50 for the GI tract is too high to pose a threat for this organ system (Table 2) (38).


The second characteristic that must be understood in order to determine if a microbe poses a human health risk is its virulence composition. Basically, virulence is the composition of armaments a microbe must have to cause disease. A microbe with high virulence means that it poses a greater threat than one with low virulence. Virulence is an overall measure of the threat of a microbe; it is composed of individual virulence factors. The classically transmissible microbes are those which have the greatest armaments. These virulence characteristics are so potent that these microbes can cause disease if even a small number reach the target organ. These include microbes such as Vibrio cholera, Salmonella typhi, Mycobacterium tuberculosis, and Group A Streptococcus, whose very names describe their diseases.

Another group of microbes, the opportunistic pathogens, is much more lightly armed. These microbes require some defect in the host (i.e., in defense mechanisms) in order to establish an infection. The defects in the host's defenses which result in infection are very specific for both the individual species of microbe and the anatomic site of the body. For example, Staphflococcus aureus is a normal inhabitant of our skin and will not cause infection under normal circumstances. If a puncture wound occurs, however, the species can be inoculated below the epidermis where the body's defense mechanisms are not efficient against it. Subsequently, a Staphylococcus aureus abscess can develop. Likewise, P. aeruginosa is commonly found in the environment. On a daily basis we ingest a large number of this species with our food, especially leafy vegetables. However, it requires very specific host defects to cause an infection. For example, only if a person experiences a severe burn over a large part of his or her body, or if his or her polymorphonuclear leukocyte count falls precipitously, can P. aeruginosa cause infection. The number of people with these profound defects are quite rare and would only be found in patients hospitalized for specific injuries or disorders (33).

The study of the virulence factors of bacteria is still in its early stages. The association with virulence has, however, been described for several specific factors (Table 3). The possession of any one factor does not mean that a microbe will cause disease. Again, like armaments, some are more powerful than others. Also, a microbe must possess more than one armament in order to breech our defenses and cause disease (Table 3) (34, 36, 37, 39, 40, 41). Dufour and Lye at the EPA conducted a survey of the ability of bacterial isolates from tap water to cause cell culture cytotoxicity (16). They found less than 2% of isolates, including Pseudomonas aeruginosa, possess cytotoxic activity. Edberg, et al., found similar results (42).

Immune Status:

The concepts of ID50 and LD50 represent the numerator of the infection equation, [number of microbes] x [virulence factors]. They define the relative risk to a normal host from a microbe.

By examining the Infection Relationship, any decrease in specific immunity will increase the likelihood of a particular organ becoming infected. Immune Status is a term that represents the body's immune systems (Table 4). Each of the body's individual immune components is separately measured (e.g., neutrophil count, lymphocyte count).

The individual components of the body's immune system - phagocytic system, humoral immunity, cellular immunity - provide specific defenses against specific microbes. Rarely is there a generalized suppression of the body's immune status. Rather, one component is decreased, thereby increasing the susceptibility of an organ system to a particular group of microbes. From drinking water, we need to be concerned with those microbes that enter the body through ingestion. For all practical purposes, drinking water microbes do not threaten to enter the body through the lungs, the skin, etc. Accordingly, in order to assign a risk to a particular microbe, one must understand which components of the gastrointestinal tract comprise the body's defense mechanisms, and which armaments of the microbes attack these particular defense mechanisms (43).

In order to specifically determine the health threat of a particular microbe from drinking water, we must address the following question: Who are the gastrointestinally immuno-suppressed? It is not correct to look in a textbook of Medical Microbiology and cross-reference a particular microbe with infection and assume the mere presence of this species in drinking water will result in acquisition of gastrointestinal infection. As the Infection Relationship demonstrates, mere contact with a microbe does not result in infection. Moreover, the only target organ at risk is the gastrointestinal tract; infections in other organs (bones, skin, etc.,) are not relevant. In fact, hospitals long ago stopped restricting bacterial contact through food and water intake. There are only two categories of patients in whom food is specially processed. These include bone marrow transplant patients and those with acute neutropenia. In both cases, the body's cellular defense mechanisms are so severely impaired that the patients have to live in sequestration (Table 4). For purposes of assessing health risk from HPC in drinking water, immunosuppression does not include people receiving immunomodulatory agents as outpatients. People receiving steroids and anti-metabolites for such conditions as rheumatoid arthritis and other connective tissue diseases are not at increased risk for gastrointestinal pathogens. Moreover, since the stomach is maintained intact, they are not at higher risk for the acquisition of gastrointestinal pathogens.

Therefore, the immune system is compartmentalized. Using Acquired Immune Deficiency Syndrome (AIDS) as an example, the immune defect is very specific. A subset of lymphocytes, known as CD4, is significantly reduced - other components of the immune system are normal. There is no evidence that AIDS patients are any more susceptible to naturally occurring HPC than persons not infected with HIV (44).

The EPA conducted an extensive review of the health effects of HPC prior to promulgating the new regulations in 1991. Furthermore, the FDA conducted a similar review in generating its new bottled water regulation. In each case the only microbiological analyte included in the regulations were the total coliform group and Escherichia coli. It was determined that total coliforms were the indicator of general water safety and E. coli the indicator of fecal contamination. Moreover, the proposed FDA regulations governing bottled water are more stringent than the EPA regulations governing tap water. The EPA regulations for tap water allow up to 5% of samples to be total coliform positive (with zero E. coli allowed) whereas the FDA has a zero tolerance limit on total coliforms in bottled water.

The fourth component in the Allen and Darby paper is that residual chlorine levels in tap water provide a level of protection not found in bottled water. As described above, if the water entering the bottle is free of pathogens, none can develop in the sealed environment provided by the bottle. In addition, although the residual chlorine in tap water will destroy the indicators of contamination (coliform, E. coli, etc.) it may not be effective at inactivating certain pathogens (protozoa, viruses, etc.) potentially associated with these indicators. Thus the absence of coliform and other indicators of contamination in tap water with residual chlorine might provide a false sense of security.

Overall Health Risk Assessment

What is the human health threat from microbes naturally found in water? First, drinking water may serve as a conduit for pathogenic microbes. For purposes of this discussion it shall be assumed that the water source and/or treatment parameters are such to preclude the presence of pathogens (e.g., Salmonella or Shigella). Moreover, it will be assumed the drinking water has been thoroughly tested for total coliforms and E. coli. Therefore, we must ask which microbes are indigenous to water and what health threat they pose.

Any place one finds water on earth one finds a natural microbial population (5, 6, 10, 11). The aqueous constituency of humans is considerably different from the aqueous niches found in the environment (Table 5). Microbes that have a lifestyle suited to the aqueous environment are not suited to maintain a continued existence in the human body. We find, therefore, that total coliforms, (excluding E. coli), such as Klebsiella and Enterobacter, and non-fermenters, such as Pseudomonas and Acinetobacter, may only intermittently be recovered from humans as part of the normal flora. Moreover, when present they are recovered in considerably smaller numbers and with less regularity than the indigenous bacteria such as E. coli (33, 34).

The vast majority of microbes isolated from drinking water do not grow well on the laboratory bacteriological culture media used for the diagnosis of human infection. Many of the isolates do not even survive propagation on these media (16). The standard bacteriological growth medium utilized to enumerate environmental bacteria (R2A agar) was designed as a low nutrient mixture to more closely mirror the environment. For example, more than 95% of the bacteria isolated from this medium will not grow when transferred to blood-containing media (16). Examination of tap and bottled water has shown that the number of bacteria tend to be rather modest (2). Rarely are there more than 50,000 HPC (equivalent to 500 to 5,000 SPC) per mL of water. Obviously, water which has been treated with chlorine or ozone tends to have a lower concentration of bacteria (6, 15).

How can we assign a risk to a particular environmental microbe? First, we must examine the numerator of the Infection Relationship. The number of microbes present in the bottle will vary markedly. There may be a large number of indigenous water bacteria present; the majority, however, will only grow on R2A agar and would not be expected to survive in the human host.

Second, based on the current state of our knowledge (16, 39, 40), it would appear that less than 2% of bacteria isolated from drinking water possess even one virulence characteristic.

Third, it is not common to find a gastrointestinal system so immunocompromised that this organ system would be susceptible to HPC. People in the community, regardless of age or therapeutics, would not be expected to have a significant decrease in immune status for microbes which are ingested. The major modulators that could change the body's susceptibility are pharmaceuticals that increase the stomach pH. The increase in pH would permit microbes that would normally die in the stomach to pass through to the intestines. While the increase in stomach pH has been associated with an increase in susceptibility to gastrointestinal pathogens such as Salmonella and Shigella, there are no data to indicate that it predisposes one to increased intestinal colonization by other microbes. Only under extraordinary circumstances, therefore, does the Immune Status component of the Infection Relationship equation becomes so decreased as to put the human host at risk from ingested microbes.

Allen and Darby provide a valuable and thought provoking review of bottled water in California. Clearly, regulations must be enforced. If not, a health threat certainly exists. However, efforts should be made to focus the limited resources of regulators and enforcers on utilizing the tools at hand to ensure the wholesomeness of water as it enters the bottle. Sampling water on the shelves for HPC or requiring expiration dates will only create confusion. Based on the components of the Infection Relationship, it is clear that bacteria found naturally in bottled water - in the absence of indicators of fecal contamination - do not pose a demonstrated health threat. Moreover, no deterioration occurs in the bottle during storage as long as the bottle remains properly sealed, thus protecting its contents from contamination.


1. Allen, L., and J.L. Darby (1994), "Quality Control of Bottled Water and Vended Water in California: A review and comparison to tap water," J. Env. Health, 56(8): 17-22.

2. Manaia, C.M., O.C. Nunes, P.V. Morais, and M.S. da Costa (1990), "Heterotrophic Plate Counts and the Isolation of Bacteria from Mineral Waters on Selective and Enrichment Media," J. Appl. Bacteriol., 69:871-876.

3. Reasoner, D.J., and E.E. Geldreich (1985), "A New Medium for the Enumeration and Subculture of Bacteria from Potable Water," Applied and Environmental Microbiology, 49:1-7.

4. Robertson, J.B., and S.C. Edberg (1993), "Technical Considerations in Extracting and Regulating Spring Water for Public Consumption," Environmental Geology, 22:52-59.

5. Reasoner, D.J. (1990), "Monitoring Heterotrophic Bacteria in Potable Water," Drinking Water Microbiology, ed. G. A. McFeters, New York: Springes-Verlag, 452-477.

6. LeChevallier, M.W., R.J. Seidler, and T.M. Evans (1980), "Enumeration and Characterization of Standard Plate Count Bacteria in Chlorinated and Raw Water Supplies," Applied and Environmental Microbiology, 40:922-930.

7. Moyer, C.L., and R.Y. Morita (1989), "Effect of Growth Rate and Starvation-Survival on the Viability and Stability of a Psychrophilic Marine Bacterium," Appl. Environ. Microbiol., 55:1122-1127.

8. Morita, R.Y. (1985), Bacteria in Their Natural Environments: The effect of nutrient conditions, Academic Press, Orlando, Fl., 111-130.

9. Roszak, D.B., and R.R. Colwell (1987), "Survival Strategies of Bacteria in the Natural Environment," Microbiol. Rev., 51:365-379.

10. LeClerc, H., D.A.A. Mossel, and C. Savage (1985), "Monitoring Non-Carbonated ('Still') Mineral Waters for Aerobic Colonization," Intern. J. Food Microbiol., 2:341-347.

11. Quevedo-Sarmiento, J., A. Ramos-Cormenzana, and J. Gonzalez-Lopez (1986), "Isolation and Characterization of Aerobic Heterotrophic Bacteria From Natural Spring Waters in Lanjaron Area (Spain)," J. Appl. Bacteriol., 61:365-372.

12. Gonzalez, C., C. Gutierrez, and T. Grande (1987), "Bacterial Flora in Bottled Uncarbonated Mineral Drinking Water," Canadian J. Microbiol., 33:1120-1125.

13. Morais, P.V., and M.S. DaCosta (1990), "Alterations in the Major Heterotrophic Bacteria Populations Isolated from a Still Bottled Mineral Water," J. Appl. Bacteriol., 69:750-757.

14. Davis, Glenn (1994), personal communication, Mi.

15. Rivilla, R., and C. Gonzalez (1988), "Simplified Methods for the Microbiological Evaluation of Bottled Natural Mineral Waters," J. Appl. Bacteriol., 64:273-278.

16. Lye, D.J., and A.P. Dufour (1991), "A Membrane Filter Procedure for Assaying Cytotoxic Activity in Heterotropic Bacteria Isolated from Drinking Water," J. Appl. Bacteriol., 70:89-94.

17. Payment, P., E. Coffin, and G. Paquette (1994), "Blood Agar to Detect Virulence Factors in Tap Water Heterotrophic Bacteria," Appl. Environ. Microbiol., 60:1179-1183.

18. Payment, P., E. Franco, L. Richardson, and J. Siemiatychi (1991), "Gastrointestinal Health Effects Associated with the Consumption of Drinking Water Produced by Point-of-use Domestic Reverse-osmosis Filtration Units," Appl. Environ. Microbiol., 57:945-948.

19. Calderon, R.L., and E.W. Wood (1988), "Bacterial Colonizing Point-of-use, Granular Activated Carbon Filters and Their Relationship to Human Health," CR-811904-01-0, U.S. Environmental Protection Agency.

20. Calderon, R.L., and E.W. Wood (1991), "Bacterial Colonizing Point-of-entry, Granular Activated Carbon Filters and Their Relationship to Human Health," CR-813978-01-0, U.S. Environmental Protection Agency.

21. Covert, T.C., E.W. Rice, S.A. Johnson, B. Bergman, C.H. Johnson, and P.J. Mason (1992)," Comparing Defined-Substrate Coliform Tests for the Detection of Escherichia coli in Water," Journal of American Water Works Association, 84:98-104.

22. Edberg, S.C., M.J. Allen, D.B. Smith, and The National Collaborative Study (1988), "National Field Evaluation of a Defined Substrate Method for the Simultaneous Enumeration of Total Coliforms and Escherichia coli from Drinking Water: Comparison with the standard multiple tube fermentation method," Appl. Environ. Microbiol., 54(6):1595-1601.

23. Edberg, S.C., M.J. Allen, D.B. Smith, and The National Collaborative Study (1989), "National Field Evaluation of a Defined Substrate Method for the Simultaneous Detection of Total Coliforms and Escherichia coli from Drinking Water: Comparison with presence-absence techniques," Appl. Environ. Microbiol., 55(4):1003-1008.

24. Federal Register (1993), "Quality Standards for Foods With No Identity Standards-Bottled Water," 58FR52042 for revision of 103.35(b).

25. Federal Register (1992), "Disinfection Requirements for Public Water Systems Which Provide Filtration," 141.72(b)(3).

26. Edberg, S.C. (1992), "Health Effects of Microbes Isolated from Drinking Water," Regulating Drinking Water Quality, ed. Charles E. Gilbert and Edward J. Calabrese, Lewis Publishers, Inc., Chelsea, Mi.

27. Guillot, E., and H. Leclerc (1993), "Biological Specificity of Bottled Natural Mineral Waters: Characterization by ribosomal ribonucleic acid gene restriction patterns," J. Appl. Bacteriol., 75:292-298.

28. Guillot, E., and H. Leclerc (1993), "Bacterial Flora in Natural Mineral Waters: Characterization by ribosomal ribonucleic acid gene restriction patterns," Syst. Appl. Microbiol., 16:483-493.

29. Edberg, S.C. (1993), Technical Assessment of the Microbiological Health Effects of Bottled Water, Yale University, New Haven, Ct.

30. Ducluzeau R., J.L. Nicolas, J.V. Galpin, and P. Raibaud (1984), "Influence de bacteries de la flore autochtone sur la survie d'Escherichia coli dans l'eau Minerale Ambouteillee," Sciences des aliments, 4:585-594.

31. Ducluzeau, R., S. Hudault, and J.V. Galpin (1977), "Influence du conditionnement sur la croissance d'une souche autochtone de Pseudomonas fluorescens dans l'eau minerale de la Grande Source de Viteel," Ann. Inst. Hydrol. Clin. XLVII, N[infinity] sp., Journees d'Hydrologie de Nancy, 77-90.

32. Ducluzeau, R. (1976), "La signification du nombre et de la nature des microorganismes telluriques presents dans l'eau minerale a l'emergence," Annli 1st. Sup. Sanit., 12:170-176.

33. Edberg, S.C. (1981), "Methods of Quantitative Microbial Analyses that Support the Diagnosis, Treatment and Prognosis of Human Infection," Critical Reviews in Microbiol., 8:339-397.

34. Duncan, I.B.R. (1988), "Waterborne Klebsiella disease," Toxicity Assessment, 3:581-598.

35. Federal Register (1973), "Quality Standards for Bottled Water-Proposed Rules," 38(226):32558-32562.

36. Brubaker, R.R. (1985), "Mechanisms of Bacterial Virulence," Annu. Rev. Microbiol., 39:21-50.

37. Finlay, B.B., and S. Falkow (1989), "Common Themes in Microbial Pathogenicity," Microbiol. Rev., 53(2):210-230.

38. Pollack, M. (1984), "The virulence of Pseudomonas aeruginosa," Rev. Infect. Dis. 6(Suppl. 3):617-626.

39. Janda, J.M., and E.J. Bottone (1981), "Pseudomonas aeruginosa Enzyme Profiling: Predictor of potential invasiveness and use as an epidemiological tool," Journal of Clinical Microbiology, 14(1):55-60.

40. Janda, J.M., S.A-Tang-Nomo, E.J. Bottone, and E.P. Desmond (1980), "Correlation of Proteolytic Activity of Pseudomonas aeruginosa with Site of Isolation," Journal of Clinical Microbiology, 12(4): 626-628.

41. Highsmith, A.K., and W. R. Jarvis (1985), "Klebsiella pneumoniae: Selected virulence factors that contribute to pathogenicity," Infect. Control, 6:75-77.

42. Edberg, S.C., P. Gallo, C. Kontnick (in press), "Analysis of the Virulence Characteristics of Bacteria Isolated from Bottled, Water Cooler, and Tap Water," Microbial Ecology in Health and Disease.

43. Duncan, H.E., and S.C. Edberg (1995), "Host-Microbe Interaction in the Gastrointestinal Tract," Critical Rev. in Microbiol., 21 (2):85-100.

44. Katlama, C., and G.M. Dickinson (1993), "Update on Opportunistic Infections," AIDS, 7(5): 185-194.

Corresponding Author: Stephen C. Edberg, Ph.D., A.B.M.M., Departments of Laboratory Medicine and Internal Medicine, Yale University School of Medicine and Clinical Microbiology Laboratory, Yale-New Haven Hospital, P.O. Box 208035, New Haven, CT 06520-8035. Telephone: (203) 785-2457; Fax: (203) 737-4170.
COPYRIGHT 1996 National Environmental Health Association
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1996, Gale Group. All rights reserved. Gale Group is a Thomson Corporation Company.

Article Details
Printer friendly Cite/link Email Feedback
Author:Edberg, Stephen C.
Publication:Journal of Environmental Health
Date:Jan 1, 1996
Previous Article:Residents in a high radon potential geographic area: their risk perception and attitude toward testing and mitigation.
Next Article:Indoor air pollution: acute adverse health effects and host susceptibility.

Related Articles
Communicating drinking water health risk: an information system that may help the process.
Drinking water quality concerns and water vending machines.
Feeding microbes to get rid of nitrates.
Gasoline additive's going, but far from gone.
Water, Water, Everywhere ...
Arsenic in Drinking Water.
Really hot water.
Manganese can make water toxic.
Toxic water is poisoning the people of Bangladesh: can a student's science project help save their lives?
Genes as pollutants: tracking drug-resistant DNA in the environment.

Terms of use | Copyright © 2017 Farlex, Inc. | Feedback | For webmasters