2014 marks the 40th anniversary of the seminal discovery by Johannes Rook, in 1974, that trihalomethanes (THMs) were formed by the chlorination of natural organic matter (NOM) in drinking water. Since this discovery, which revolutionized how we viewed drinking water safety and quality, hundreds of other classes of disinfection by-products (DBPs) have been discovered. The finding in 1976 by the US National Cancer Institute that chloroform, the dominant THM, was a rodent carcinogen spurred a large number of epidemiology and toxicology studies into chlorinated drinking water. In 1985, this cancer finding was shown to be wrong. We should now be asking: What do we know about the human health impacts of DBPs in drinking water? Bladder cancer has been the most consistent finding from epidemiologic studies in North America and Europe and the possibility that chlorinated drinking water contributes an increased risk of bladder cancer remains a viable hypothesis. Despite some recent improvements in exposure assessments to focus on inhalation and dermal exposures rather than ingestion, no causal agent with sufficient carcinogenic potency has been identified, nor has a mechanistic model been validated. Consequently, a sensible precautionary approach to managing DBPs remains the only viable option based on four decades of evidence.

LIST OF ABBREVIATIONS

     
  • BDCM

    bromodichloromethane

  •  
  • CH

    choral hydrate

  •  
  • CxDBP

    chlorination disinfection by-product

  •  
  • DBAN

    dibromoacetonitrile

  •  
  • DBCM

    dibromochloromethane

  •  
  • DBP

    disinfection by-product

  •  
  • DCAA

    dichloroacetic acid

  •  
  • DCAN

    dichloroacetonitrile

  •  
  • HAA5

    sum of five haloacetic acids, MCAA, DCAA, TCAA, monobromoacetic acid (MBAA) and dibromoacetic acid (DBAA)

  •  
  • MCAA

    monochloroacetic acid

  •  
  • NDMA

    N-nitrosodimethylamine

  •  
  • NOM

    natural organic matter

  •  
  • TBM

    tribromomethane, bromoform

  •  
  • TCAA

    trichloroacetic acid

  •  
  • TCP

    trichlorophenol

  •  
  • THM

    trihalomethane

  •  
  • THM4

    sum of chloroform, BDCM, DBCM and TBM

INTRODUCTION

The year 2014 provided a major anniversary in the history of drinking water quality and safety assessment. In 1974, the Dutch water chemist, Rook (1974) published his seminal discovery that trihalogenated methanes (THMs) are formed by the reaction of chlorine used to disinfect drinking water (inactivate pathogenic microorganisms) and natural organic matter (NOM). This discovery was soon followed by the publication of Bellar et al. (1974) who independently made the same discovery in the USA. This single discovery forever changed how we look at drinking water quality and has led to the subsequent discovery of hundreds of other unintended by-products of essentially all disinfection processes. These contaminants are now broadly termed, disinfection by-products (DBPs).

There have likely been many thousands of research papers published on various aspects of DBPs and drinking water (e.g., a June 2014 Web of Science search turned up over 12,000 citations for only a few of the relevant key words), countless graduate students trained and certainly several billions of dollars of investment in technology, monitoring and research over the past 40 years. There have clearly been major, undeniable benefits of all this research activity and investment in yielding improved understanding of drinking water treatment processes, better drinking water quality in terms of both safety and aesthetic quality and an overall higher level of evidence-based understanding for the provision of safe drinking water. However we should be willing to ask how well the past 40 years of research has improved our understanding of the nature and extent of any human health impacts associated with DBPs in drinking water.

This paper seeks to provide a high level overview of some of the major issues about chlorination DBPs and human health over the 40 year timeline before reflecting on our current state of knowledge and what can reasonably be anticipated in the coming decade. This focus on chlorination DBPs has been chosen because chlorine-based disinfection processes are still the most widely practiced and these are generally the most cost-effective for smaller systems that are resource-constrained. Our focus in this overview discussion of 40 years of research and regulation of DBPs necessarily emphasizes toxicology evidence and risk assessment rather than epidemiology. This emphasis is necessary because epidemiologic studies of drinking water are not able to establish mechanisms of toxic effect attributable to specific DBPs that invariably occur as complex mixtures. Likewise, it is outside our scope to engage in a detailed comparison and critique of the strengths and weaknesses of toxicology vs. epidemiology.

MATERIAL AND METHODS

This discussion is directed only at organic DBPs and emphasizes those produced by disinfection processes involving chlorine. Such DBPs may or may not contain chlorine (e.g., NDMA) and for the purposes of this discussion will be termed chlorination disinfection by-products (CxDBPs). The discussion is based on a distillation of the perspectives gained from performing a series of comprehensive literature reviews over the past decade (Hrudey 2008, 2009), participation on several international expert panels (e.g. Mills et al. 1998; Arbuckle et al. 2002; Sinclair et al. 2002; Hrudey et al. 2003; Hrudey et al. 2012, International Interdisciplinary Expert Panel 2015) on issues relating to DBPs and human health over the past 17 years, and co-editing (authoring or co-authoring several chapters) for the book – Disinfection By-Products and Human Health (Hrudey & Charrois 2012).

RESULTS AND DISCUSSION

The short format allowed for these papers does not allow for comprehensive or detailed review of this subject, but more detailed discussions may be found elsewhere (Hrudey 2008, 2009, 2012a; International Interdisciplinary Expert Panel 2015). The following relies on a few illustrative examples to demonstrate the larger issues.

Initial concerns about adverse health effects possibly attributable to drinking water exposure to DBPs initially focused on cancer. This was initially driven by the coincidence that chloroform was declared a carcinogen in 1976 by the US National Cancer Institute (NCI 1976). Less than 10 years later, evidence was published (Jorgenson et al. 1985) to demonstrate that the original rodent tumour results were an artefact of the bioassay experimental protocol caused by delivering near-lethal doses of chloroform dissolved in corn oil as a daily bolus (by gavage) into the animal gut leading to a massive liver tumour response. The later work delivering extremely high doses (900,000 μg/L rats, 1,800,000 μg/L mice) of chloroform to the rodents via drinking water produced no detectable tumour response. Related research established that chloroform is not genotoxic (does not damage DNA) and that the original bioassay results were the result of massive localized tissue damage caused by the high chloroform exposure delivered by the corn oil vehicle. More than 10 years later, the US EPA proposed in 1998 to recognize this evidence and to treat chloroform as a threshold carcinogen (i.e. a substance capable of causing tumours at exposures only above a specified level; in such cases there is no validity to using a cancer slope factor (CSF) to estimate cancer risk), but this proposal was withdrawn until a court judgement obtained by the chlorine industry ordered the US EPA to follow the requirements of the US Safe Drinking Water Act (SDWA), i.e. to use the best available science (Pontius 2000). Ironically, at trial the US EPA acknowledged the threshold mode of action. The US regulatory changes regarding chloroform as a carcinogen were proposed in 2003 and finally adopted in 2006.

The increasing WHO guideline numbers for chloroform (30 μg/L in 1984, then 200 μg/L in 1993, then 300 μg/L in 2011; Table 1), reflect the changing understanding (declining concern) for chloroform in drinking water as a cancer risk. The numbers for THM provide a different story even though chloroform is the dominant THM in most waters, unless bromide levels cause substantial formation of brominated THMs.

Table 1

Drinking water guidelines for organic disinfection by-products (DBPs)a

DatebJurisdiction or agencyInitiativeComment
1978 Health Canada THM4 350 μg/L MAC, guideline, not to exceed 
1979 US EPA, Safe Drinking Water Act (SDWA) THM4 100 μg/L Maximum contaminant level (MCL), regulation, running annual average 
1984 World Health Organization (WHO) Chloroform 30 μg/L MAC, guideline, not to exceed 
1993 WHO Chloroformc 200 μg/L MAC, guideline, not to exceed 
TBM 100 μg/L MAC, guideline, not to exceed 
DBCM 100 μg/L MAC, guideline, not to exceed 
BDCM 60 μg/L MAC, guideline, not to exceed 
TCP 200 μg/L MAC, guideline, not to exceed 
1996 Australia, National Health & Medical Research Council (NHMRC) THM4 250 μg/L MAC, guideline, not to exceed 
MCAA 150 μg/L MAC, guideline, not to exceed 
DCAA 100 μg/L MAC, guideline, not to exceed 
TCAA 100 μg/L MAC, guideline, not to exceed 
CH 100 μg/L (2014) MAC, guideline, not to exceed 
1996 Health Canada THM4 100 μg/L MAC, guideline, running annual average, quarterly samples 
1998 US EPA, SDWA THM4 80 μg/L MCL, regulation, system-wide running annual average, quarterly samples 
HAA5 60 μg/L 
1998 EC Drinking water directive THM4 100 μg/L MCL at the point of delivery as a maximum in each sample. Requirement to achieve lower if possible by improving treatment. 
2003 to 2004 WHO DCAA 50 μg/L MAC, guideline, not to exceed 
DCAN 20 μg/L MAC, guideline, not to exceed 
DBAN 70 μg/L MAC, guideline, not to exceed 
MCAA 20 μg/L MAC, guideline, not to exceed 
TCAA 200 μg/Ld MAC, guideline, not to exceed 
2006 US EPA, SDWA THM4 80 μg/L MCL, regulation, locational running annual average, quarterly samples - to include locations with highest concentrations 
HAA5 60 μg/L 
2006 Health Canada BDCM 16 μg/L MAC, guideline, not to exceed 
2006 WHO NDMA 0.10 μg/L MAC, guideline, not to exceed 
2008 Health Canada THAA (HAA5) 80 μg/L MAC, guideline, running annual average, quarterly samples 
2009 Health Canada BDCM withdrawne New NTP cancer bioassay for BDCM 
2011 Health Canada NDMA 0.04 μg/L MAC, guideline, not to exceed 
2011 Australia, NHMRC NDMA 0.10 μg/L MAC, guideline, not to exceed 
2011 WHO chloroform 300 μg/Lf,g MAC, guideline, not to exceed 
DatebJurisdiction or agencyInitiativeComment
1978 Health Canada THM4 350 μg/L MAC, guideline, not to exceed 
1979 US EPA, Safe Drinking Water Act (SDWA) THM4 100 μg/L Maximum contaminant level (MCL), regulation, running annual average 
1984 World Health Organization (WHO) Chloroform 30 μg/L MAC, guideline, not to exceed 
1993 WHO Chloroformc 200 μg/L MAC, guideline, not to exceed 
TBM 100 μg/L MAC, guideline, not to exceed 
DBCM 100 μg/L MAC, guideline, not to exceed 
BDCM 60 μg/L MAC, guideline, not to exceed 
TCP 200 μg/L MAC, guideline, not to exceed 
1996 Australia, National Health & Medical Research Council (NHMRC) THM4 250 μg/L MAC, guideline, not to exceed 
MCAA 150 μg/L MAC, guideline, not to exceed 
DCAA 100 μg/L MAC, guideline, not to exceed 
TCAA 100 μg/L MAC, guideline, not to exceed 
CH 100 μg/L (2014) MAC, guideline, not to exceed 
1996 Health Canada THM4 100 μg/L MAC, guideline, running annual average, quarterly samples 
1998 US EPA, SDWA THM4 80 μg/L MCL, regulation, system-wide running annual average, quarterly samples 
HAA5 60 μg/L 
1998 EC Drinking water directive THM4 100 μg/L MCL at the point of delivery as a maximum in each sample. Requirement to achieve lower if possible by improving treatment. 
2003 to 2004 WHO DCAA 50 μg/L MAC, guideline, not to exceed 
DCAN 20 μg/L MAC, guideline, not to exceed 
DBAN 70 μg/L MAC, guideline, not to exceed 
MCAA 20 μg/L MAC, guideline, not to exceed 
TCAA 200 μg/Ld MAC, guideline, not to exceed 
2006 US EPA, SDWA THM4 80 μg/L MCL, regulation, locational running annual average, quarterly samples - to include locations with highest concentrations 
HAA5 60 μg/L 
2006 Health Canada BDCM 16 μg/L MAC, guideline, not to exceed 
2006 WHO NDMA 0.10 μg/L MAC, guideline, not to exceed 
2008 Health Canada THAA (HAA5) 80 μg/L MAC, guideline, running annual average, quarterly samples 
2009 Health Canada BDCM withdrawne New NTP cancer bioassay for BDCM 
2011 Health Canada NDMA 0.04 μg/L MAC, guideline, not to exceed 
2011 Australia, NHMRC NDMA 0.10 μg/L MAC, guideline, not to exceed 
2011 WHO chloroform 300 μg/Lf,g MAC, guideline, not to exceed 

aAll abbreviations are explained in list of abbreviations before the Introduction.

bEarliest date adopted. Later dates are only shown if the number has changed with later updates.

cThe recognition that chloroform was not genotoxic led to a threshold cancer risk assessment.

dMAC value increased because TCAA weight of evidence is judged not to be genotoxic.

eThe result of a 2006 National Toxicology Program (NTP) bioassay on BDCM removed the basis for the risk calculation by finding no evidence of cancer risk for BDCM administered in drinking water.

fMAC value increased because of a larger proportion of exposure assigned to water, noting that other uses of chloroform have decreased substantially since the previous MAC set in 1993.

gWHO states: ‘Disinfection should not be compromised in attempting to control disinfection byproducts’ (WHO 2011).

Another interesting development evident in Table 1 was the Health Canada guideline for bromodichloromethane (BDCM) in 2006 that was withdrawn in 2009. Our expert panel (Hrudey et al. 2003) had suggested that Health Canada should take a closer look at BDCM because it potentially posed a reproductive risk concern, but the 2006 maximum acceptable concentration (MAC) of 16 μg/L was calculated using a CSF for a cancer bioassay done on BDCM that also suffered from using the corn oil daily bolus dose regimen. When a new study for BDCM was published (NTP 2006) using drinking water, BDCM-dosing failed to support the rodent cancer risk at concentrations up to 700,000 μg/L, Health Canada to its credit, convened an expert panel to review its BDCM guideline and retracted it.

The rationale for maintaining THM4 levels much lower than harmless levels of chloroform is not articulated clearly by drinking water agencies. Presumably it reflects caution about brominated THM species, where they occur, because they may pose a greater health risk (WHO 2005; Health Canada 2009). Perhaps it reflects the concern that within the overall large number of CxDBPs there may be issues and reducing these using THMs as a surrogate is desirable, particularly within the bounds of achievability. Unfortunately, this lack of clarity has led to considerable confusion among water practitioners about whether chloroform, as the dominant THM, poses a cancer risk via drinking water exposure (Hrudey 2012b). Several audience surveys were conducted in the past 3 years, with varying degrees of formality and structure with operational personnel in Australia and Canada to determine their understanding about whether THM regulatory levels are based on cancer risk posed by chloroform. Responses ranged from over 50% to as high as 80% who believe that THM levels in drinking water are set to protect against cancer risk from chloroform, with the second largest response being ‘don't know’. Only a small minority of respondents knew that chloroform in drinking water does not pose any cancer risk. Who can blame the majority for believing in a chloroform cancer risk in the absence of any clear communications on this issue by regulators?

Certainly, the research literature is not helpful in providing clarity for readers on this topic. Chowdhury & Hall (2010) published predictions of cancer cases for Canadian cities that claimed to use a CSF for chloroform from the US EPA IRIS database (checked in 2009) except that the CSF had been removed from IRIS in 2001. Ultimately, the Chowdhury & Hall (2010) paper was retracted by the journal, but another paper (Chowdhury et al. 2011), was published in another journal using similar erroneous cancer case calculations for each Canadian province. This one avoided claiming use of a CSF from the US EPA IRIS database, instead the journal editors allowed the authors to use a CSF from an unspecified source, describing it only as a ‘previously published value.’ The 2011 paper made the additional error of presenting CSF-based risk predictions (which are 70-year lifetime risks) as being annual cancer risk predictions. This paper attracted a commentary exchange (Bull et al. 2012) to reveal its errors. Drinking water professionals busy with the job of producing quality drinking water cannot be expected to keep track of such exchanges and should question the quality of the research literature peer review process for such cases.

Despite the confusion about chloroform, there is a viable hypothesis that CxDBPs in drinking water may cause human bladder cancer (Hrudey 2012a; International Interdisciplinary Expert Panel 2015). This possibility is based on bladder cancer showing the greatest consistency as a cancer outcome that is statistically associated with human exposure to chlorinated drinking water in up to 10 epidemiological studies of varying predictive quality and statistical power. Because of this possibility, the case for continuing research and sensible precautionary management of CxDBPs in drinking water clearly remains justified, particularly the removal of precursor NOM which should reduce most CxDBPs. However, this precautionary rationale needs to be effectively articulated by regulators and other authorities.

The site-specific (bladder) cancer epidemiology risk estimates are orders of magnitude greater than the upper bound, non-site specific cancer risk prediction for any known carcinogenic CxDBP. Some will argue that since CxDBPs are always present as a mixture there could be synergistic interactions. However, at such low concentrations there is little evidence that this is possible and the most likely outcome would be additive effects from substances with similar mechanisms of action or even target organs. However, in view of the small concentrations even simple additivity is not likely to result in the massive increase in toxic effects that would be required to match the epidemiological risk estimates.

Bull (2012) explored the evidence about bladder cancer for those DBPs that have sufficient evidence to have a published CSF available. Not even the sum of risk from all the carcinogenic CxDBPs with CSFs, yields a total cancer risk prediction remotely (within 100 fold) close to the magnitude of the cancer risk indicated by the bladder cancer epidemiology. Bull et al. (2011) have explored this discrepancy by considering other classes of DBPs that are more plausible carcinogens, but so far even considering these new DBPs fails to close the gap between toxicological risk assessment estimates and epidemiologic estimates for bladder cancer that might be caused by DBPs.

The scope of this short paper does not allow for adequately exploring the evidence in support of a causal relationship between exposure to CxDBPs and adverse reproductive outcomes. However, as a generalization, the human evidence on reproductive impacts of CxDBPs is less consistent than it is for a possible causal association with bladder cancer (Nieuwenhuijsen et al. 2009, 2010).

Figure 1 provides a rationale to address the reality that contaminants detected in drinking water are not equally risky to human health. Even judging concentrations of any substance detected in relation to its established guideline or regulatory value does not provide the entire meaningful perspective. As evident in Table 1, different jurisdictions have judged the same health research evidence on a DBP to yield different criteria. But, more important, not all contaminants that could pose a health risk in drinking water at sufficiently high concentration have done so, except possibly in extreme spill or direct poisoning episodes, or the comparative confidence that is evident for high levels of drinking water arsenic causing cancer (IARC 2004).

Figure 1

Disinfection by-products assessed according to a rationale showing two key dimensions of any risk of a drinking water contaminant causing any serious adverse human health effect (adapted from Hrudey et al. 2012).

Figure 1

Disinfection by-products assessed according to a rationale showing two key dimensions of any risk of a drinking water contaminant causing any serious adverse human health effect (adapted from Hrudey et al. 2012).

For the specific case of DBPs, the possible human health risks that have been suggested in some cases (e.g., bladder cancer) have apparent risk magnitudes that certainly warrant attention, but the level of confidence which can be placed in the evidence for actual human risks via drinking water is vastly lower than the established certainty that microbial pathogens cause human illness via drinking water exposure. Only microbial pathogens and a relatively short list of chemical contaminants like arsenic, fluoride and possibly nitrates/nitrites and selenium have caused documented cases of human illness via drinking water exposure (WHO 2007).

An issue that seems to influence common perceptions of the CxDBP and human health issue regards an inability for some to distinguish between mere detection of a substance and the existence of a human dose of that substance sufficiently high to cause a health effect.

Figure 2 is a simple graphical illustration of the toxicologists’ ‘creed’: the dose makes the poison – all substances are toxic in sufficient dose using the widely experienced toxic substance, ethanol. About 400 mL of pure ethanol would be lethal to a person of average body mass. Toxic effects of ethanol occur for daily consumption of concentrations above about 1023 molecules/L but we are mostly able to tolerate and repair this ‘damage’. However, the completely harmless level of ethanol in fresh, pure orange juice (mean ∼400 ppm or 5 × 1021 molecules/L, Nisperos-Carriedo & Shaw 1990) is 100 fold lower and concentrations down to a million times lower can be detected by routine analysis. Given sufficient investment, much lower concentrations could be found. In principle, as little as 1 molecule (100) could be detected in a litre of water. Hopefully, no one is likely to defend a case that detection of 1 molecule of ethanol per litre of water poses a health risk. The key point that many of our colleagues will say they recognize but which they may not truly believe is that mere detectability of a substance tells us nothing about likely human health impacts of negligibly small exposures.

Figure 2

The dose makes the poison – an illustration, best read from left to right, of the enormous range of concentrations (requiring a logarithmic scale). Concentrations are presented as molecules per litre (shown as powers of 10) of substances found in drinking water using ethanol, a well-documented toxic substance, as the example.

Figure 2

The dose makes the poison – an illustration, best read from left to right, of the enormous range of concentrations (requiring a logarithmic scale). Concentrations are presented as molecules per litre (shown as powers of 10) of substances found in drinking water using ethanol, a well-documented toxic substance, as the example.

Only knowing how much exposure to a toxic substance occurs relative to the amount proven necessary to cause a human toxic effect can tell us anything meaningful about the likelihood of adverse human health impacts.

The past 40 years of experience with DBPs have demonstrated that our collective ability to detect substances in water has totally outstripped our capability to judge the human health risk posed. We can be certain that the lower end of the range of newly detected DBPs will continue toward the right of Figure 2. Currently, we are only half way down the scale. We have already moved detectability of water contaminants about 7 decades (10,000,000 fold) down that scale in the past 50 years. Completing the journey to detecting one molecule per litre will not likely take another 50 years.

CONCLUSIONS

Overall, the discovery of DBPs in drinking water has been positive in driving research to improve drinking water quality and optimize disinfection and other treatment processes because of the implied risk trade-off between DBPs and pathogens, but as with any situation involving balancing of risk, there has been continuous tension.

There will be inevitable discovery of more DBPs in drinking water, but this reality need not pose any human health risk contrary to a common ‘popular’ view implicitly held by too many of our colleagues; simple detection of any in drinking water provides no evidence about the likelihood of causing human health effects. The real challenge is to find ways to improve our ability to judge whether any of the CxDBPs that are detectable in drinking water pose any human health risk. Concluding that CxDBPs must cause human health effects because they are ‘toxic’ substances, is untenable. Authentic evidence concerning human health risks for the longest known DBPs remains uncertain. However, the dominant CxDBP, chloroform, essentially poses no cancer risk in drinking water at levels below about a few hundred μg/L.

Moving forward, a sensible, precautionary approach is necessary, whereby currently known technologies to manage DBPs are applied as part of overall water quality management systems. However, extraordinary measures to reduce DBPs or abandon chlorine-based disinfection are currently not justified. Likewise, arbitrary measures to reduce one or more classes of DBP must not be implemented without full understanding of what different DBPs or other water quality problems may result from technology change. Above all, management of DBPs must remain a much lesser concern than the essential requirement of assuring that drinking water is adequately and consistently disinfected at all times.

REFERENCES

REFERENCES
Arbuckle
T. E.
Hrudey
S. E.
Krasner
S. W.
Nuckols
J. R.
Richardson
S. D.
Singer
P.
Mendola
P.
Dodds
L.
Weisel
C.
Ashley
D. L.
Froese
K. L.
Pegram
R. A.
Schulz
I. R.
Reif
J.
Bachand
A. M.
Benoit
F. M.
Lynberg
F. M.
Poole
C.
Waller
K.
2002
Assessing exposure in epidemiologic studies to disinfection by-products in drinking water: Report from an International Workshop
.
Environmental Health Perspectives
110
(
Suppl. 1
),
53
60
.
Bellar
T. A.
Lichtenberg
J. J.
Kroner
R. C.
1974
The occurence of organohalides in chlorinated drinking waters
.
Journal American Water Works Association
66
(
12
),
703
706
.
Bull
R. J.
2012
Toxicological evaluation of experimental data that informs the magnitude of cancer risk from disinfection by-products
. In:
Disinfection By-Products and Human Health
(
Hrudey
S. E.
Charrois
J. W. A.
, eds).
IWA Publishing
,
London
.
Bull
R. J.
Fawell
J.
Cotruvo
J.
Hrudey
S. E.
2012
Critique of Chowdhury et al. (2011)
. Journal of Hazardous Materials
237–238
,
382
383
.
Chowdhury
S.
Hall
K.
2010
RETRACTED
.
Human health risk assessment from exposure to trihalomethanes in Canadian cities
.
Environment International
36
,
453
460
.
Chowdhury
S.
Rodriguez
M. J.
Sadiq
R.
2011
Disinfection byproducts in Canadian provinces: associated cancer risks and medical expenses
.
Journal of Hazardous Materials
187
,
574
584
.
Health Canada
2009
Guidelines for Canadian Drinking Water Quality:Guideline Technical Document. Trihalomethanes. (May 2006 with April 2009 addendum). http://hc-sc.gc.ca/ewh-semt/pubs/water-eau/trihalomethanes/index-eng.php.
Hrudey
S. E.
2008
Chlorination disinfection by-products in drinking water: A primer for public health practitioners reviewing evidence from over 30 years of research
.
National Collaborating Centre for Environmental Health
.
Hrudey
S. E.
2012a
Epidemiological inference and evidence on DBPs and human health. Chapter 11. 213–282
. In:
Disinfection By-Products and Human Health
(
Hrudey
S. E.
Charrois
J. W. A.
, eds).
IWA Publishing
,
London
.
Hrudey
S. E.
2012b
Disinfection By-Products and Human Health. Watermark, B.C
.
Water and Wastewater Association
21
(
2
),
30
31
.
Hrudey
S. E.
Charrois
J. W. A.
(eds)
2012
Disinfection By-Products and Human Health
.
IWA Publishing
,
London
.
May 2012
.
324
pp.
Hrudey
S. E.
Bathija
A.
Bull
R. J.
Dodds
L.
Jenkins
G.
King
W.
Murphy
P.
Narotsky
M.
Reif
J.
Savitz
D.
Thomas
B.
2003
Executive Summary. International Expert Workshop for Health Canada to Identify Critical Endpoints for Assessment of Health Risks Related to Trihalomethanes (THMs) in Drinking Water. March 2003
.
Hrudey
S. E.
Fawell
J.
Leiss
W.
Rose
J. B.
Sinclair
M.
2012
Managing Uncertainty in the Provision of Safe Drinking Water
.
Canadian Municipal Water Consortium, Canadian Water Network. University of Waterloo
.
IARC
2004
Some Drinking Water Disinfectants and Contaminants, Including Arsenic. IARC Monographs on the Evaluation of Carcinogenic Risks to Humans. Vol 84
.
IARC Scientific Publications, International Agency for Research on Cancer
,
Lyon
,
France
.
International Interdisciplinary Expert Panel
2015
Evidence for Association of Human Bladder Cancer with Chlorination Disinfection By-Products
.
Water Research Foundation
,
Denver, CO, USA
.
Jorgenson
T. A.
Meierhenry
J. R.
Rushbook
C. J.
Bull
R. J.
Robinson
M.
1985
Carcinogenicity of chloroform in drinking water to male Osborne-Mendel rats and female B6C3F1 mice
.
Fundamental & Applied Toxicology
5
,
760
769
.
Mills
C. J.
Bull
R.
Cantor
K.
Reif
J.
Hrudey
S. E.
Huston
P.
&
an Expert Working Group
1998
Health risks of drinking water chlorination byproducts: report of an expert working group
.
Chronic Diseases in Canada
19
(
3
),
91
102
.
NCI
1976
Carcinogenesis Bioassay of Chloroform
.
National Cancer Institute
,
Bethesda, MD
,
USA
.
Nieuwenhuijsen
M. J.
Grellier
J.
Smith
R.
Iszatt
N.
Bennett
J.
Best
N.
Toledano
M.
2009
The epidemiology and possible mechanisms of disinfection by-products in drinking water
.
Philosophical Transactions of the Royal Society A.
367
,
4043
4076
.
Nieuwenhuijsen
M. J.
Grellier
J.
Iszatt
N.
Martinez
D.
Rahman
M. B.
Villanueva
C. M.
2010
Literature review of meta-analyses and pooled analyses of disinfection by-products in drinking water and cancer and reproductive health outcomes. Chapter 24, 483–496
. In:
Contaminants of Emerging Concern in the Environment: Ecological and Human Health Considerations
(
Halden
R. U.
, ed.).
ACS Symposium Series, American Chemical Society
.
Nisperos-Carriedo
M. O.
Shaw
P. E.
1990
Comparison of volatile flavor components in fresh and processed orange juices
.
Journal of Agricultural and Food Chemistry
38
(
4
),
1048
1052
.
NTP
2006
Toxicology and Carcinogenesis Studies of Bromodichloromethane (CAS No. 75–27–4) in Male F344/N Rats and Female B6C3F1 Mice (Drinking Water Studies)
.
National Toxicology Program
,
Research Triangle Park, NC, USA
.
Pontius
F. W.
2000
Chloroform: science, policy and politics
.
Journal American Water Works Association
92
(
5
),
12, 14, 16, 18
.
Rook
J. J.
1974
Formation of haloforms during chlorination of natural waters
.
Water Treatment and Examination
23
,
234
243
.
Sinclair
M.
Bull
R.
Froese
K.
Hrudey
S. E.
Nieuwenhuijsen
M.
Rizak
S.
2002
Disinfection By-Products and Health Effects. Summary of an International Expert Workshop
.
CRC for Water Quality and Treatment
.
Occasional Paper No. 5
.
Salisbury
,
South Australia
.
WHO
2005
Trihalomethanes in Drinking-water. Background document for development of WHO Guidelines for Drinking-water Quality. World Health Organization. http://www.who.int/water_sanitation_health/dwq/chemicals/THM200605.pdf.
WHO
2007
Chemical safety of drinking-water: Assessing priorities for risk management
.
World Health Organization
.
WHO
2011
Guidelines for Drinking-water Quality
,
4th edn. World Health Organization
.