Workshop Report
Health Risks of Drinking Water Chlorination By-products: Report of an Expert Working Group

Christina J Mills, Richard J Bull, Kenneth P Cantor, John Reif, Steve E Hrudey, Patricia Huston and an Expert Working Group

Introduction

A number of recent epidemiologic studies, including a 1995 study sponsored by Health Canada, have found a modest increase in the risk of bladder cancer among people who had drinking water that included high levels of chlorination by-products. Other studies of water chlorination by-products have suggested possible increased risks of colon and rectal cancers, as well as adverse reproductive and developmental effects, such as increased spontaneous abortion rates and fetal anomalies.

Chlorination by-products are created as a result of water purification procedures that have been used for decades to prevent the spread of microbial disease. Chlorination has been hailed as one of the most important public health initiatives of the century. Thus, any examination of the need for further action regarding the human health risks from chlorination by-products must not compromise microbial disinfection. In Canada, the currently acceptable level of the most common by-products, the trihalomethanes (THMs), is 100 µg/L. Other disinfectants, such as chloramine and ozone, also create by-products. The toxicity of these by-products has not been extensively studied.

These concerns led the Laboratory Centre for Disease Control to question whether current Canadian policies on chlorination by-products should be re-examined in light of the evolving body of evidence on their risks. Subsequently, a meeting was held in Ottawa on May 1-2, 1997, with leading epidemiologists, toxicologists, public health specialists and water quality experts. The meeting's objectives were to obtain authoritative advice on cancer and reproductive risks associated with exposure to water chlorination by-products, to determine their likely importance for public health and to advise Health Canada on how to proceed.

A three-step process was undertaken to meet the objectives. Participants were given background papers and a set of key questions to review prior to the meeting. At the workshop, experts presented an overview of what was known to date on water chlorination by-products from toxicologic studies, epidemiologic studies of cancer and adverse reproductive/developmental effects, and risk assessment. Following this, participants responded to key questions and arrived at conclusions and recommendations.

This paper summarizes the information provided in the background papers and presentations, describes the consensus arrived at regarding assessment of evidence for level of risk and presents suggestions for future research.

Toxicology

Richard J Bull

Chlorination of drinking water is the most cost-effective means to prevent the spread of waterborne infections and has been a common public health method for almost a century. In 1974, a major class of chlorination by-products, the trihalomethanes (THMs), was identified as occurring in much higher concentrations in chlorinated water than in source water. THMs are produced from the interaction of chlorine with naturally occurring organic materials.

The many by-products produced by water chlorination have been classified broadly as halogenated or non-halogenated by-products (Table 1). The most commonly occurring halogenated by-products are the THMs; within this group of compounds, chloroform is the by-product found most frequently and at the highest concentrations. A second commonly occurring class of by-products is the haloacetic acids, which include dichloroacetic acid (DCA) and trichloroacetic acid (TCA). Non-halogenated by-products are generally natural substrates or metabolites.

The major determinant of by-product concentration is the level of organic material in the source water. For this reason, water facilities that derive their water from surface waters (lakes, rivers, reservoirs) produce water with higher levels of by-products than facilities that draw from ground waters (wells, springs). After chlorination, THM concentrations range from 30 to 150 µg/L in surface water, and from 1 to 10 ##181;g/L in ground waters. The type and quantity of by-products formed is determined by the amount and character of organic material, as well as the ambient pH and bromide concentration in the water.

Animal Studies of Carcinogenesis

A comprehensive toxicologic assessment of chlorination by-products has been difficult due to the many by-products involved and the different modes of action that may result in carcinogenesis. To date, animal studies have focused on by-products with the greatest human exposure or toxicologic concern (Table 2). As a result, the most frequent tumour type observed was liver cancer; this was found in mice and rats after exposure to THMs as well as haloacetates.

The mechanism of cancer induction appears to vary with different by-products and different species. Chloroform, for example, seems to cause cancer by a non-genotoxic (or epigenetic) mechanism and only after massive exposure. Some cancers were species-specific; for example, trichloroacetate produced liver cancer in mice, but not rats. Furthermore, liver cancer from chlorinated by-products has never been found in humans. This suggests that by-products cause liver cancer either through mechanisms that are species-specific or from exposure levels that are much higher than current standards.

Some of the rarer THMs-such as bromodichloro- methane-induce colon cancer in mice. Dibromoacetate has been associated with the development of aberrant crypt foci in the distal colon of rats. These findings are of particular interest because colon cancer has been associated with exposure to high levels of THMs in some epidemiologic studies.

Animal Studies of Developmental and Reproductive Effects

Most of the toxicologic research on chlorination by-products has focused on carcinogenesis, but in light of recent epidemiologic data, studies of developmental and reproductive effects merit review (Table 3). The most consistent developmental finding was soft tissue abnormalities, including ventricular septal defects. Exposure to haloacetonitriles was associated with embryo death in rats. Degeneration of testicular epithelium has been found in rats and dogs after exposure to haloacetates, but no correlate has been found in human studies.

Although animal evidence demonstrates that high levels of by-product exposure induce cancer in laboratory animals, a number of intriguing issues remain. No single chlorinated by-product studied in toxicologic studies appears to be carcinogenic at human levels of exposure. Furthermore, evidence for carcinogenesis differs between toxicologic and epidemiologic studies: by-product exposure is most commonly associated with liver cancer in animals and bladder cancer in humans. These differences raise concerns about the appropriateness of current cancer risk estimates derived from animal studies.

It is now recognized that risks from chlorinated drinking water cannot be determined accurately by simply summing up the toxicologic hazards of each individual by-product. Initial toxicologic studies of by-product mixtures have produced little convincing evidence of adverse effect, but this cannot be extrapolated to humans in part because of the diversity in by-product mixtures in treated water currently available. Future research will need to be hypothesis-driven to address this complex issue.

Cancer Epidemiology

Kenneth P Cantor

Over the last 20 years, considerable epidemiologic research has examined possible associations between cancer and water chlorination by-products. The quality of the research has improved greatly over this period, to the point that it is now debatable whether to include earlier studies in a critical overview. The first epidemiologic studies were ecologic, correlating age-adjusted sex- and race-specific regional cancer mortality rates with reported chlorinated surface water supplies versus chlorinated or non-chlorinated well water supplies. The sites of cancers most frequently associated with chlorinated water are bladder, colon and rectum.

The results of earlier studies stimulated a generation of case-control studies using mortality records to identify cases and comparison groups. In most of these studies, the water supply was determined by the last place of residence, as noted on the death certificate. Some used birth place (also recorded on the death certificate) or obtained water exposure histories from interviews with next of kin.

In 1992, Morris and colleagues published a meta-analysis to assess the evidence for a relationship between chlorination of drinking water and neoplastic disease. Ten studies were included in the final analysis. Using statistics provided in each study and a random effects model, the researchers derived a single estimate of relative risk for each organ-specific neoplasm.

The meta-analysis found that exposure to chlorinated surface water was associated with a statistically significant increased relative risk of bladder cancer (odds ratio [OR] = 1.20) and rectal cancer (OR = 1.34). Controlling for available confounding variables, such as smoking, urban living and occupation, did not diminish the risks. The estimated risk for colon cancer was not statistically significant, but incidence increased proportionally with dose.

There were multiple problems with these studies. Many relied on rough estimates of by-product exposure, measurement of confounding variables was inconsistent and some studies suffered from selection bias and poor response rates. Studies have now been conducted that include more accurate exposure data and track additional potential confounding factors, which gives their results more weight.

Tables 4, 5 and 6 highlight these improved epidemiologic studies. Relative risks are inferred from calculations of odds ratios in most studies. For simplicity, we present a single relative risk to summarize a rich and complex body of data. A result greater than 1.0 is interpreted as a positive risk; less than 1.0, as a negative risk. Relative risks are interpreted as "statistically significant" if their associated 95% confidence intervals do not include 1.0 and "not statistically significant" if they do.

Colon Cancer

Table 4 summarizes nine studies assessing the risk of colon cancer after exposure to chlorinated water by-products. Among the seven earlier studies, two showed a significantly positive result. Inconsistent findings emerged from the two most recent studies (Marrett and King [1995] and Hildesheim [1998]), both case-control investigations of newly diagnosed disease.

Marrett and King studied over 5000 people in Ontario; approximately 950 had bladder, colon or rectal cancer. Age- and sex-matched controls were identified from the general population. THM levels were estimated back to 1950 in regional water supplies, using a survey of water treatment facility history and measurements of THM. People with exposure to THMs greater than or equal to 50 µg/L for more than 35 years were 1.5 times more likely to develop colon cancer, and the data demonstrated a dose-response relationship that persisted after accounting for potential confounding factors such as nutrient, caloric and fibre intake.

Hildesheim and colleagues conducted a study in Iowa, identifying 685 colon cancer patients. The control group consisted of 2400 people matched for age, sex and having developed one of five other types of cancer. Estimates of exposure to THM and to chlorinated surface water were made for the full lifetime of all subjects and adjustments were made for confounding variables.

In contrast to the Marrett and King study, Hildesheim et al. found no elevated risk of colon cancer. While the methods to estimate THM exposure were somewhat more precise in the Marrett and King study, it is unlikely that this would explain the absence of an association in the Hildesheim study. These contradictory findings are not currently understood. They may be due to chance, to water quality differences between Ontario and Iowa or to other factors.

In conclusion, the evidence for an increased risk of colon cancer from exposure to chlorination by-products is inconclusive.

Rectal Cancer

Table 5 identifies eight studies that address the possible association of rectal cancer with chlorination by-products. Of the six earliest studies, two showed a statistically significant risk of cancer associated with by-product exposure. Once again, the two most recent studies had inconsistent results: the Marrett and King study showed no association, whereas the Hildesheim study showed a statistically significant positive association and a positive duration-response relationship.

In summary, the evidence for an association between rectal cancer and chlorinated by-products is also inconclusive. However, in light of the positive finding in the meta-analysis, the evidence is somewhat stronger for rectal cancer than colon cancer.

Bladder Cancer

Evidence of a link between chlorination by-products and bladder cancer is more consistent than it is for colon and rectal cancers. Table 6 outlines 11 studies that assessed the association of bladder cancer with THM exposure. Three of seven studies published prior to 1990 were statistically significant. King and Marrett's 1996 article reported a relative risk of 1.61 for exposure to an estimated THM level of 50 µg/L or greater for 35 years or more. Excess risk was found only after more than 20 years of exposure, and risk increased with time. Results suggested an increased risk with higher concentrations of by-products and (counter-intuitively) a lower risk in smokers; neither of these trends were statistically significant.

McGeehin et al. (1993) conducted a cancer registry-based study that identified a control group from patients with cancer other than bowel or bladder cancer in order to eliminate recall bias. Their findings were similar to the previous studies: long-term exposure to chlorinated water increased the relative risk of bladder cancer by 1.8. Unlike the King and Marrett study, cigarette smoking was strongly associated with an increased risk of bladder cancer.

After assessing the different variables, McGeehin et al. found that THM was no longer a statistically significant predictor of risk (although years of chlorinated water consumption was) if they took out the 1989 THM concentrations. This suggests that THM may be a surrogate marker rather than the causal agent, or it may simply reflect a statistical design artifact.

Cantor and colleagues conducted two studies in this area. One (1987) was a large case-control study with 3000 cases and 6000 controls. Unfortunately, only half the population came from places with enough regional variability in water supply characteristics to conduct a meaningful analysis. These cases revealed a 1.8 relative risk of bladder cancer in those who had consumed water with high THM levels over a long period of time. This was the same for men and women; the association was stronger for non-smokers than smokers.

More recently, Cantor et al. (1998) looked at 1450 bladder cancer cases and 2400 controls in Iowa. They gathered a lifetime residential history, information on other risk factors and estimated THM levels. Bladder cancer risk among current or previous smokers with long-term exposure to chlorination by-products was about twice the risk of smokers who had not been exposed to chlorinated water. A recent study by Freedman (1997) found similar results.

In summary, there were five epidemiologic studies that showed a statistically significant positive association of chlorinated by-product exposure with risk of bladder cancer. King and Marrett (1996) estimated that 14-16% of bladder cancers may be attributable to chlorinated water. Our understanding of this phenomenon, however, remains limited by the fact that all studies relied on retrospective exposure estimates.

Reproductive and Developmental Effects

John Reif

The evidence for reproductive and developmental effects associated with exposure to chlorination by-products is scant. Only five studies on this topic have been published; several others are pending. Most published studies use a case-control method and rely on birth certificates and birth defect registries; all lack important individual data.

If there are true adverse reproductive outcomes due to exposure to chlorination by-products during pregnancy, they should be more readily detectable than true carcinogenic effects because gestation offers a short latent period for by-product exposure.

Spontaneous Abortion, Stillbirth and Pre-term Delivery

Table 7 summarizes findings on the risk of spontaneous abortion, stillbirth and pre-term delivery after exposure to chlorination by-products. Only one study (Savitz [1995]) looked at spontaneous abortion rates. This hospital-based, case-control study included exposure assessment based on interviews and data from five public water supplies. Several confounding variables were taken into account: maternal age, poverty level, smoking and alcohol history. The relative risk of miscarriage in those women exposed to greater levels of chlorination by-products was slightly increased, but was not statistically significant.

Two large studies examined risk of stillbirth after exposure to chlorination by-products. Aschengrau (1993) conducted a hospital-based, case-control study of over 14,000 pregnancies. Exposure assessment was based on the municipal water supply of the mother's place of residence at the time of the pregnancy outcome, and the usual confounding variables were measured. Researchers found a 2.6-fold increase in risk for stillbirth in those exposed to chlorinated surface water, but this was not statistically significant.

The largest study to date, by Bove (1992; 1995), included over 80,000 births and almost 600 fetal deaths. The study revealed a negative correlation between stillbirth and exposure to chlorination by-products, once again not statistically significant.

Four studies looked at pre-term delivery. In one that was population-based, Kramer et al. (1992) used information from birth certificates to identify the water supply. Each case was matched to five controls and exposure measures included all THMs. After adjusting for age, parity, education, smoking and prenatal care, the researchers found no increased risk of prematurity among those who were exposed to higher THM levels during pregnancy than those who were not.

Kanitz et al. (1996) compared pre-term delivery rates between two towns with similar social and economic characteristics and the same perinatal care services, but different water supplies. One town had chlorinated water with sodium hypochlorite and chlorine dioxide, the other had untreated well water. After adjusting for the usual confounding variables, researchers observed a small increase in risk of prematurity in newborns of mothers who drank the water treated with chlorine dioxide.

Low Birth Weight and Growth Retardation

Table 8 consolidates the findings on low birth weight and growth retardation linked to exposure to chlorination by-products. Four studies examined low birth weight; all found some increase in risk, but only one showed statistically significant risk.

The studies by Bove (1992; 1995) and Kramer (1992) were the only two that looked at growth retardation (small-for-gestational age). Both showed a small and statistically significant increase in risk.

Birth Defects

Some preliminary evidence suggests that exposure to chlorination by-products during pregnancy is associated with birth defects (Table 9). Examining the records of offspring of women exposed to chlorination by-products during pregnancy, Bove (1995) observed a significantly increased risk of total anomalies; neural tube and oral cleft defects were the most common. An increased risk of cardiac defects also appeared (consistent with animal studies), although this was not statistically significant.

Epidemiologic research in the area of reproductive and developmental effects is still at an early stage. The studies to date are inadequate to infer causality. However, on the basis of available studies, there is evidence to suggest a weak association between chlorination by-products and adverse fetal growth and moderate evidence for an association with congenital malformations.

Risk Assessment of Chlorination By-products

Steve E Hrudey

Risk assessment interprets available evidence in a formalized way in order to inform regulatory decision making. For example, risk assessment of chlorination by-products resulted in the development of maximum acceptable concentrations (MACs) as noted in Health Canada's Guidelines for Canadian Drinking Water Quality.

The currently acceptable level of THMs (100 µg/L) has been calculated to carry a lifetime cancer risk of less than 4 X 10^-6-an essentially negligible risk. To understand how the acceptable level was derived at, one needs to understand classic cancer risk assessment. Once this is known, the difficulties of applying this to the case of chlorinated by-products will become apparent.

Cancer Risk Assessment

There are two key questions in cancer risk assessment: "How likely is a particular agent to be a human carcinogen?" and "If it is one, what are the cancer risks for a given exposure scenario?"

The classic approach to risk assessment was established in 1983 by the National Academy of Sciences and includes four stages.

• Hazard identification documents previous evidence to determine what adverse outcomes a substance may cause. It should provide some answers to the first question on likelihood.
• Dose-response assessment summarizes quantitative data about the dose of a substance and observed adverse outcomes. It usually involves extrapolation of observations from high level exposures in animal studies in order to make predictions for much lower level exposures in humans.
• Exposure assessment documents and estimates actual human exposures. This is necessary to determine what doses should be used with the dose-response assessment.
• Risk characterization synthesizes the information from the first three stages into a quantitative expression of the risk to hypothetical individuals or populations. (This involves two different approaches, depending on whether the substance is a genotoxic carcinogen or not.) This final step should answer the second question on cancer risks.

Acceptable exposure levels are usually established at doses much lower than experimental levels. Two methods have been used to set such environmental criteria.

• The "No observed adverse effect level" (NOAEL) uncertainty factor approach assumes a threshold for a dose-response curve. The method depends upon defining the highest dose at which no adverse effect can be observed in animal studies. A tolerable daily intake or reference dose is then calculated by dividing the NOAEL by a product of uncertainty and modifying factors. These uncertainty factors include other considerations such as differences between experimental animals and humans, variability in individual sensitivity and quality of evidence.
• The Risk Specific Dose approach typically uses an upper bound estimate on risk derived from a low dose linear extrapolation (through zero dose) derived from rodent bioassays. Most of these studies have been conducted with only two or three doses, the maximum tolerated dose (MTD) and fixed fractions of the MTD. The MTD has been defined as the highest dose that causes no more than a 10% weight loss, no excess mortality, no clinical signs of toxicity and no unexpected pathologic lesions. Recent concerns about the meaning of the model predictions have led to alternate proposals, such as the Tumorigenic Dose 05 (TD05), proposed by Meek and Long (1996).

Toxicologic and epidemiologic data are usually both included in risk assessments, but in the case of chlorination by-products they have identified different outcomes. Further research is needed to clarify these differences in order to provide better evidence on which to formulate appropriate water quality levels.

Consensus of the Expert Working Group

After hearing the presentations of the evidence, the Expert Working Group deliberated the two questions below and arrived at the following consensual opinions.

1. Given currently available evidence, how likely is it that chlorination by-products cause cancer/reproductive effects in humans? If likely (possible or probable), how important a public health problem is it?

Cancer

The Working Group noted that the evidence for this must be reviewed on a site-specific basis. Participants concluded that it was possible (60% of the group) to probable (40%) that chlorination by-products pose a significant risk to the development of cancer, particularly bladder cancer.

The risk of bladder cancer, and possibly other cancers, poses a risk to public health; this is a moderately important public health problem.

Reproductive and Developmental Effects

There is presently insufficient evidence to establish a causal relationship between exposure to chlorinated by-products and adverse reproductive outcomes in humans. However, further research is warranted.

If the suggested findings of the limited data are confirmed, chlorinated by-products in current surface water supplies could pose an important health problem. Even a relatively small excess risk of adverse reproductive outcomes associated with chlorinated by-products may contribute to a large absolute number of adverse outcomes since these outcomes are quite common: 10-20% of all pregnancies terminate in spontaneous abortion; birth defects occur in 1-2% of all live births.

2. Given the state of the current evidence, are there enough quantitative data to be useful in an in-depth quantitative risk/benefit/cost evaluation?

There are not enough quantitative data at this time to conduct an in-depth quantitative risk/benefit/cost evaluation. However, a mechanism to monitor the human health risks, such as an Expert Working Group, should be established to advise Health Canada and make recommendations as to when the evidence has accumulated to the point that a more in-depth evaluation is warranted.

Recommendations Regarding Future Research

Throughout the meeting, there were suggestions from participants regarding research needs and priorities. These suggestions, summarized in Table 10 according to risk factor categorization for thematic consistency, were not systematically discussed or approved by the Expert Working Group as a whole and do not represent a consensus on research priorities.

Bibliography

This bibliography consists of references provided by the experts who presented evidence to the Expert Working Group.

Toxicology

Bull RJ, Sanchez IM, Nelson MA, Larson JL, Lansing AL. Liver tumour induction in B6C3F1 mice by dichloroacetate and trichloroacetate. Toxicology 1990;63:341-59.

Bull RJ, Meier JR, Robinson M, Ringhand HP, Laurie RP, Stober JA. Evaluation of the mutagenic and carcinogenic properties of brominated and chlorinated acetonitriles: by-products of chlorination. Fundam Appl Toxicol 1985;5:1065-75.

Butterworth BE, Conolly RB, Morgan KT. A strategy for establishing mode of action of chemical carcinogens as a guide for approaches to risk assessments. Cancer Letts 1995;93:129-46.

Cicmanec JL, Condie LW, Olson GR, Wang SR. 90-day toxicity study of dichloroacetate in dogs. Fundam Appl Toxicol 1991;17:376-89.

Daniel FB, DeAngelo AB, Stober JA, Olson GR, Page NP. Hepato-carcinogenicity of chloral hydrate, 2-chloroacetaldehyde, and dichloroacetic acid in the male B6C3F1 mouse. Fundam Appl Toxicol 1992;19:159-68.

DeAngelo AB, Daniel FB, Stober JA, Olson GR. The carcinogenicity of dichloroacetic acid in the male B6C3F1mouse. Fundam Appl Toxicol 1991;16:337-47.

DeAngelo AB, Daniel FB, Most BM, Olson GR. The carcinogenicity of dichloroacetic acid in the male Fischer 344 rat. Toxicology 1996;114:207-21.

Dunnick JK, Haseman JK, Likja HS, WyandS. Toxicity and carcinogenicity of chlorodibromo-methane in Fischer 344/N rats and B6C3F1 mice. Fundam Appl Toxicol 1985;5:1128-36.

Epstein DL, Nolen GA, Randall JL, Christ SA, Read EJ, Stober JA, Smith MK. Cardiopathic effects of dichloroacetate in the fetal Long-Evans rat. Teratology 1992;46:225-35.

Herren-Freund SL, Pereira MA, Khoury MD, Olson G. The carcinogenicity of trichloroethylene and its metabolites, trichloroacetic acid and dichloroacetic acid, in mouse liver. Toxicol Appl Pharmacol 1987;90:183-9.

Jorgenson TE, Meierhenry EF, Rushbrook CJ, Bull RJ, Robinson J. Carcinogenicity of chloroform in drinking water to male Osborne-Mendel rats and female B6C3F1 mice. Fundam Appl Toxicol 1985;5:760-9.

Lander RE, Klinefelter GR, Strafer LF, Cereous JD, Dyer CJ. Acute spermatogenic effects of bromoacetic acids. Fundam Appl Toxicol 1994;22:422-30.

Lander RE, Klinefelter GR, Strafer LF, Cereous JD, Roberts NL, Dyer CJ. Spermatotoxicity of dibromoacetic acid in rats after 14 daily exposures. Reprod Toxicol 1994;8:251-9.

National Cancer Institute. Report on the carcinogenesis bioassay of chloroform. Bethesda (MD): National Cancer Institute, 1976; NTIS PB-264-018.

National Toxicology Program. Technical report on the toxicology and carcinogenesis studies of chlorodibromomethane. US Department of Health and Human Services 1984; Publ 84-2583.

National Toxicology Program. Toxicology and carcinogenesis studies of bromodichloromethane in F344/N rats and B6C3F1 mice (gavage studies). US Department of Health and Human Services, 1987; Technical Report Series No 321.

National Toxicology Program (NTP). Toxicology and carcinogenesis studies of tribromomethane (bromoform) in F344/N rats and B6C3F1 mice (gavage studies). US Department of Health and Human Services, 1989; Technical Report Series No 350.

Pereira MA. Carcinogenic activity of dichloroacetic acid and trichloroacetic acid in the liver of female B6C3F1 mice. Fundam Appl Toxicol 1996;31:192-9.

Smith MK, Randall JL, Stober JA, Read EJ. Developmental toxicity of dichloracetonitrile: a by-product of drinking water disinfection. Fundam Appl Toxicol 1989;12:765-72.

Smith MK, Randall JL, Tocco DR, York RG, Stober JA, Read EJ. Teratogenic effects of trichloroacetonitrile in the Long-Evans rat. Teratology 1988;113-20.

So B-J, Bull RJ. Dibromoacetate (DBA) acts as a promoter of abnormal crypt foci in the colon of F344 rats. Toxicologist 1995;15:1242.

Toth GP, Kelty KC, George EL, Read EJ, Smith MK. Adverse male reproductive effects following subchronic exposure of rats to sodium dichloroacetate. Fundam Appl Toxicol 1992;19:57-63.

Cancer Epidemiology

Alvanja M, Goldstein I, Susser M. A case-control study of gastrointestinal and urinary tract cancer mortality and drinking water chlorination. In: Jolley RL, Gorchev H, Hamilton DH Jr, editors. Water chlorination: environmental impact and health effects. 2nd ed. Ann Arbor (MI): Ann Arbor Science Publishers, 1978:395-409.

Brenniman GR, Vasilomanolakis-Lagos J, Amsel J, Tsukasa N, Wolff AH. Case-control study of cancer deaths in Illinois communities served by chlorinated or non-chlorinated water. In: Jolley RL, Brungs WA, Cumming RB, et al, editors. Water chlorination: environmental impact and health effects. 3rd ed. Ann Arbor (MI): Ann Arbor Science Publishers, 1980:1043-57.

Cantor KP, Lynch CF, Hildesheim M, Dosemeci M, Lubin J, Alavanja M, Craun GF. Drinking water source and chlorination by-products in Iowa: 1. Risk of bladder cancer. Epidemiology 1998;9:21-8.

Cantor KP, Hoover R, Harge P, Mason TJ, Silverman DT, Altman R, et al. Bladder cancer, drinking water source and tap water consumption: a case-control study. J Natl Cancer Inst 1987;79:1269-79.

Cragle DL, Shy CM, Struba RJ, Siff EJ. A case-control study of colon cancer and water chlorination in North Carolina. In: Jolley RL, editor. Water chlorination chemistry, environmental impact and health effects. Chelsea (MI): Lewis Publishers, 1985:153-9.

Freedman DM, Cantor KP, Lee NL, Chen LS, Lei HH, Ruhl CE, et al. Bladder cancer and drinking water: a population-based case-control study in Washington County, Maryland. Cancer Causes Control 1997;8:738-44.

Gottlieb MS, Carr JK, Clarkson JR. Drinking water and cancer in Louisiana: a retrospective mortality study. Am J Epidemiol 1982;116:652-67.

Hildesheim ME, Cantor KP, Lynch CF, Dosemeci M, Lubin J, Alavanja M, et al. Drinking water source and chlorination byproducts in Iowa: II. Risk of colon and rectal cancers. Epidemiology 1998;9:29-35.

King WD, Marrett LD. Case-control study of bladder cancer and chlorination by-products in treated water (Ontario, Canada). Cancer Causes Control 1996;7:596-604.

Marrett LD, King WD. Great Lakes Basin cancer risk assessment: a case-control study of cancers of the bladder, colon, and rectum. Ottawa: Bureau of Chronic Disease Epidemiology, Health Canada; 1995 Jul.

McGeehin MA, Reif JS, Becher J, Mangione EJ. A case-control study of bladder cancer and water disinfection methods in Colorado. Am J Epidemiol 1993;138:492-501.

Wilkins JR, Comstock GW. Source of drinking water at home and site-specific cancer incidence in Washington County, Maryland. Am J Epidemiol 1981;114:178-90.

Young TB, Wolf DA, Kanarek MS. Case-control study of colon cancer and drinking water trihalomethanes in Wisconsin. Int J Epidemiol 1987;16:190-7.

Young TB, Wolf DA, Kanarek MS. Epidemiologic study of drinking water chlorination and Wisconsin female cancer mortality. J Natl Cancer Inst 1981;67:1191-8.

Zierler S, Feingold RA, Danley RA, Craun G. Bladder cancer in Massachusets related to chlorinated and chloraminated drinking water: a case-control study. Arch Environ Health 1988;43(2):195-200.

Zierler S, Danley RA, Feingold L. Types of disinfectant in drinking water and patterns of mortality in Massachusetts. Environ Health Perspect 1986;69:275-79.

Reproductive and Developmental Effects

Aschengrau A, Zierler S, Cohen A. Quality of community drinking water and the occurrence of late adverse pregnancy outcomes. Arch Environ Health 1993;48:105-13.

Bove FJ, Fulcomer MC, Klotz JB, Esmart J, Dufficy EM, Savrin JE. Public drinking water contamination and birth outcomes. Am J Epidemiol 1995;141:850-62.

Bove FJ, Fulcomer MC, Klotz JB, Esmart J, Dufficy EM, Zagraniski RT. Report on Phase IV-A: public drinking water contamination and birthweight, fetal deaths, and birth defects. A cross-sectional study. Trenton (NJ): New Jersey Dept of Health, 1992 Apr.

Kanitz, et al. Association between drinking water disinfection and somatic parameters at birth. Environ Health Perspect 1996;104(5):516-20.

Kramer MD, Lynch CF, Isacson P, Honson JW. The association of waterborne chloroform with intrauterine growth retardation. Epidemiology 1992;5:407-13.

Reif JS, Hatch MC, Bracken M, Holmes LB, Schwetz BA, Singer PC. Reproductive and developmental effects on disinfection byproducts in drinking water. Environ Health Perspect 1996;104:1056-61.

Savitz DA, Andrews KW, Pastore LM. Drinking water and pregnancy outcome in central North Carolina: source, amount and trihalomethane levels. Environ Health Perspect 1995;103:592-6.

Risk Assessment

Bull RJ, Kopfler FC. Health effects of disinfectants and disinfection byproducts. Denver: American Water Works Association Research Foundation, 1991.

Butterworth, et al. In: Bull RJ, et al. Water chlorination: essential process or cancer hazard? Fundam Appl Toxicol 1995;28:155-66.

Dourson, et al. Evolution of science-based uncertainty factors in non-cancer risk assessment. Regul Toxicol Pharmacol (in press).

Health Canada. Guidelines for Canadian drinking water quality. 6th ed. Ottawa, 1996; Cat H48-10/1996E.

Krewski D, Thomas RD. Carcinogenic mixtures. Risk Anal 1992;12:105-13.

Meek ME, Long G. Health-based tolerable daily intakes/ concentrations and tumorigenic doses/ concentrations for priority substances. Ottawa, 1996; Health Canada Cat H46-2/96-194E.

National Research Council. Risk assessment in the federal government: managing the process. Washington (DC): National Academy Press, 1983.

Expert Working Group

Tye Arbuckle* (Health Canada); Frank Bove (US Agency for Toxic Substances and Disease Registry); Richard Bull (Battelle Pacific Northwest Laboratories); Byron E Butterworth (Chemical Industry Institute of Toxicology); Kenneth P Cantor* (US National Cancer Institute); Anthony B DeAngelo (US Environmental Protection Agency); Fred Hauchman (US Environmental Protection Agency); Steve E Hrudey (University of Alberta); Kenneth Johnson* (Health Canada); Will King (Queen's University); Gary Klinefelter (US Environmental Protection Agency); Patrick Levallois* (Centre de Santé publique de Québec); Ernest Mastromatteo (Canadian Public Health Association); Mike McGeehin* (US Centers for Disease Control and Prevention); Rekha Mehta (Health Canada); Christina Mills* (Health Canada); Joel Paterson* (Health Canada); John Reif (Colorado State University); Jack Siemiatycki, Chair (International Agency for Research on Cancer, Institut Armand-Frappier); Robert G Tardiff (The Sapphire Group); Barry Thomas* (Health Canada).

* Member of organizing committee  

[Table of Contents] [Next]