International Agency for Research on Cancer (IARC) - Summaries & Evaluations

(Group 3)

For definition of Groups, see Preamble Evaluation.

VOL.: 52 (1991) (p. 45)

5. Summary of Data Reported and Evaluation

5.1 Exposure data

Water supplies were first chlorinated at the turn of the century, and over the following two decades chlorination was introduced for disinfection of drinking-water in most industrialized countries. In the chlorination process, chlorine reacts mainly with natural water constituents to produce a complex mixture of by-products, including a wide variety of halogenated compounds, the actual levels of which depend on the amount of chlorine added and the type of water source. In general, groundwaters produce lower levels, while surface waters often tend to produce higher levels of chlorination by-products; however, there is some evidence that groundwaters can give higher levels of brominated substances, probably due to higher levels of bromide in the untreated water. Estimates of the total halogenated organic matter generated during chlorination suggest typical levels in the range < 10-250 mg/l as chlorine. The main chlorination by-products are trihalomethanes and chlorinated acetic acids, which usually occur in the range 1-100 mg/l (although higher levels have been reported). Many products occur in the range 1-10 mg/l, while a large number can be detected at levels of < 1 mg/l. The by-products responsible for most of the bacterial mutagenicity found in chlorinated drinking-water, 3-chloro-4-(dichloromethyl)-5-hydroxy-2[5H]- furanone (MX) and associated substances, are present at very low concentrations (< 0.1 mg/l).

5.2 Experimental carcinogenicity data

Two series of studies were considered to provide evidence that could support an evaluation of the potential carcinogenicity of chlorinated drinking-water.

Samples of material concentrated from treated and undisinfected or treated and chlorinated water samples were tested in mice in three initiation-promotion experiments (by subcutaneous injection followed by topical application of 12-O-tetradecanoylphorbol 13-acetate). None of the concentrates derived from the chlorinated water induced a significantly increased incidence of skin tumours when compared with concentrates derived from undisinfected water samples or with saline.

In one experiment in mice, oral administration of chlorinated humic acids in the drinking-water did not increase the incidence of tumours over that in animals receiving unchlorinated humic acids or in saline-treated controls.

5.3 Human carcinogenicity data

Seven case-control studies conducted in the USA were considered to provide evidence that could support an evaluation. Four of these had community exposure data, and three had individually derived exposure data. The four studies with community exposure data each included several cancer sites. One study showed a significant increase in risk for colon cancer only; another showed a significant increase only for rectal cancer; the other two studies showed no excess risk for cancer.

Of the three case-control studies with individual exposure data, one was a population-based study of urinary bladder cancer carried out by interview in 10 areas of the USA. Many potential confounding factors, including smoking, were taken into account in the analyses. An early analysis of the study showed a significant association between long-term use at home of a chlorinated surface water source (as compared to an unchlorinated groundwater source) and urinary bladder cancer in nonsmokers only. In a subsequent analysis, tap-water intake was considered in addition to home water source, and consumption level of tap water was significantly associated with urinary bladder cancer; this effect was substantially confined to those who had lived for 40 years or more in a house with a chlorinated surface water source. There were significant and increasing trends in urinary bladder cancer risk with duration of residence in a house with a chlorinated surface water source for both women and nonsmokers whose tap-water consumption was above the median. In a further report based only on Iowa participants in this study, risk for urinary bladder cancer was associated with duration of use of a chlorinated water source, and the association became stronger with increasing accuracy of the exposure measure.

In the second of these case-control studies, carried out in Massachusetts (USA), the authors reported an excess risk for mortality from urinary bladder cancer among people who had lived in areas with chlorinated water supplies as compared with chloraminated supplies. Some confounding factors, including smoking, were taken into account; however, the proportion of eligible subjects for whom exposure could be ascertained was low.

In a third case-control study, based on deaths among members of the New York State Teachers' Retirement System, no association was found between deaths from cancers of the colon and rectum combined and estimated use of surface water or intake of chloroform from domestic and workplace water supplies over the 20 years prior to death. Few confounding variables were taken into acount.

A cohort of the general population in a county in Maryland (USA) was enrolled and surveyed in 1963 and followed up to 12 years. Urinary bladder cancer incidence was found to be higher in both men and women residents supplied mainly by a chlorinated surface water source compared with county residents who obtained their drinking-water from unchlorinated deep wells; but the effects of chlorinated drinking-water could not be distinguished from factors related to urbanicity, and the numbers were too small to rule out a chance effect.

Six correlation studies and one time-trend study were considered by the Working Group to provide some useful data. These studies showed moderately consistent patterns of a positive correlation between use of surface water or of chlorinated water and cancers of the stomach, colon, rectum, urinary bladder and lung, with the most consistent patterns for cancers of the urinary bladder and rectum.

The studies that were considered informative, and therefore included in this summary, were nevertheless difficult to interpret in an evaluation of the carcinogenicity of chlorinated drinking-water. The water variables studied - whether surface or groundwater and others - were generally imperfect surrogates for the subject of this monograph. There is cause for some scepticism about the estimates of exposure to chlorinated drinking-water in all of these studies. Furthermore, very few attempted to document exposure over long periods of the subjects' lives. Chlorination by-products differ according to local conditions and practices of chlorination, and the health effects found in one place may not be found elsewhere. Many variables, such as smoking habits, dietary practices and environmental conditions, influence the risks for cancer, and they may differ between populations served by chlorinated and unchlorinated water supplies. Such factors should ideally be taken into account in an epidemiological study; however, in most of the studies evaluated, there was little if any information available about them. When the data are examined on the basis of individual cancer sites, the evidence of elevated risk is strongest for cancer of the urinary bladder. The strongest study of cancer at this site supports the hypothesis of an elevated risk due to drinking chlorinated surface water compared with unchlorinated groundwater. However, the sum of the evidence from other studies, although showing some degree of consistency, is severely compromised by the weaknesses outlined above.

5.4 Other relevant data

Elevated serum cholesterol levels were reported in women but not in men living in communities served by chlorinated versus nonchlorinated water supplies in one study. No difference in the prevalence of ancephaly was observed between villages served by chlorinated and nonchlorinated groundwater in another study.

In regard to studies of genetic and related effects, only those reports were included in which the role of chlorination could be evaluated. Samples of unconcentrated chlorinated drinking-water were not genotoxic in bacteria or in a micronucleus assay in plants and did not induce morphological transformation in cultured mammalian cells. Samples of organic material concentrated from chlorinated surface waters were usually genotoxic in bacteria and induced sister chromatid exchange, micronuclei and chromosomal aberrations in single studies with cultured mammalian cells. In a single study, no activity was observed in a mammalian cell assay for mutation.

Samples of organic material concentrated from chlorinated groundwaters were less frequently mutagenic in bacteria than those from chlorinated surface waters; in a single study, they induced sister chromatid exchange but not micronuclei in cultured mammalian cells.

Samples of organic material concentrated from surface water treated with either chlorine dioxide or ozone followed by chlorination induced mutation in bacteria in some studies.

5.5 Evaluation

There is inadequate evidence for the carcinogenicity of chlorinated drinking-water in humans.

There is inadequate evidence for the carcinogenicity of chlorinated drinking-water in experimental animals.

Overall evaluation

Chlorinated drinking-water is not classifiable as to its carcinogenicity to humans (Group 3).

For definition of the italicized terms, see Preamble Evaluation.

Last updated: 17 November 1997

    See Also:
       Toxicological Abbreviations