Drinking water disinfection, for example by chlorine, has decreased the number of waterborne diseases dramatically. In the last thirty years the potential health risk of chemical disinfectants for drinking water has gained a lot of attention. Since the discovery of chloroform in chlorinated drinking water, several epidemiological and laboratory studies have been carried out. These studies show that chloroform is carcinogenic for laboratory animals when exposure to large doses takes place.

What are the research themes?

Many disinfection byproducts are bio accumulative. They are not destroyed by the body and can accumulate in body tissues. Research on health effects of disinfection byproducts aims at the following themes:

- Health effects on humans that drink disinfected drinking water. The research is carried out through epidemic studies. These are mostly concerned with long-term effects. Humans are exposed to small concentrations of disinfection byproducts for many years.

-Toxicity of separate disinfection byproducts and mixtures of disinfection byproducts. This research is carried out on laboratory animals.

What research is carried out on laboratory animals?

The research on laboratory animals comes across several difficulties. There is a great number of disinfection byproducts. Cancer can develop in different ways. Various laboratory animals react differently on disinfection byproducts. Research on laboratory animals aims mainly on disinfection byproducts with the highest human exposure rate and the highest toxicity rate.

Rats are used for research of health effects of disinfection byproducts

Do disinfection byproducts affect development and reproduction of laboratory animals?

Most researches carried out on reproduction effects of disinfection byproducts aim at birth defects and spontaneous abortion. Little research has been carried out on effects on male reproduction. An American research shows that bromodichloromethane (BDCM) and chloral hydrate (CH) lower the speed and mobility of sperm in laboratory rats. The effect of BDCM at low concentrations is stronger than the effect of CH or other disinfection byproducts that lower sperm speed. (Klinefelter, 1996)

Research of the effect of a combination of disinfection byproducts on rats

In 2002 American researchers looked at the carcinogenity of a mixture of disinfection byproducts on rats. Aim of this research was to see whether the carcinogenity of different disinfection byproducts would be increased in a mixture and what the effect was of long-term exposure to low concentrations disinfection byproducts. Eker rats, which are sensitive to effects of renal carcinogenic substances, were used. Both male and female rats were exposed to disinfection byproducts dichloromethylhydroxyfuranone (MX), potassium bromate (KBrO₃), chloroform (CHCl₃) and bromodichloromethane (BDCM) in drinking water during 4 to 10 months. These disinfection byproducts were chosen, because they are proven renal carcinogenic or poisonous to kidneys. Both low and high concentrations were used. Mixtures of disinfection byproducts and separate disinfection byproducts were used. There seemed to be a relation between dosage and the origination of renal cancer. There was no difference in the amount of tumors in kidneys, uterus or spleen between the mixture of disinfection byproducts or the disinfection byproduct with the largest effect. This research shows that a mixture of these disinfection byproducts does not have a higher risk of cancer than the disinfectant with the greatest effect (Hooth, 2002)

Effects of chlorine dioxide, chlorite and chlorate on laboratory animals

A survey of available research results on acute and long time exposure of rats, mice, chickens to chlorine dioxide, chlorate and chlorite in drinking water show that all these animals experienced alterations in their blood cells. These effects were dose related and occurred only at high concentrations (up to 1g/L). With these long time studies rats were supplied with chlorine dioxide concentrations up to 1 g/l and sodium chlorite or -chlorate concentrations up to 100 mg/L during 30 or 60 days. At chlorite concentrations of 100 mg/l or more, the amount of red blood cells and haemoglobin decreased. After exposure of 90 days these effects decreased. The results for mice were comparable. 
Furthermore chlorine dioxide, chlorite and chlorate alter DNA in testes and kidneys. This may imply that these substances have effects on reproduction. The outcome of these researches can not directly be transferred to people. More research has to be carried out. (Couri, 1982)

Chloroform

Chloroform, a disinfection byproduct of chlorine, is one of the most investigated trihalomethanes. Toxicological research (Larson, 1994a) shows that chloroform causes damage to liver and finally causes cancer when it is daily directly applied into the stomach of laboratory animals. The amount of chloroform is too big for the liver to break down completely. The liver is damaged and death of cells and regenerative cell growth occur. The risk on cell mutation and cancer in exposed organs is increased. Another research was carried out in which laboratory animals were exposed to the same amount of chloroform dissolved in drinking water. They did not develop cancer. This was probably due to the fact that throughout the day animals were exposed to small amounts of drinking water with chloroform. The liver was able to break down the chloroform without getting damaged. (Larson et al., 1994b) Butterworth et al., 1998). The Environmental Protection Agency (EPA) concludes that as long as exposure to chloroform remains under given threshold values that cause cell damage, the risk for cancer is very low. Standards set for chloroform in drinking water are far below these values. (EPA, 1998)

Did research of health effects of disinfection byproducts on humans take place?

Besides experiments with laboratory animals (rats and mice) there are also epidemical studies on effects of exposure of humans to disinfection byproducts in drinking water. At first the relation of death because of cancer and the use of chlorinated or non-chlorinated drinking water were investigated. 
Subsequently studies showed that chlorinated drinking water increases the risk on bladder and anal cancer. The risk did not decrease when other factors, smoking, residence and work were investigated as well. The risk of intestinal cancer was not significant, but increased at higher concentrations disinfection byproducts. (Morris , 1992) 

At first the relationship of drinking chlorinated water and cancer.

The health effects of halogenated disinfection byproducts can be divided in two groups. They can be carcinogenic or have effects on reproduction and development:

Are disinfection byproducts carcinogenic?

In the 1960's new methods, gas chromatography and mass spectrometry, were developed to identify chemical substances at very low concentrations. Awareness of the large amount and diversity of chemical substances in air and water was born. In 1974 EPA enacted a list of 187 organic substances that are found in drinking water. Some of these substances are carcinogenic or mutagenic. Only a few substances, including trihalomethanes chloroform, bromoform, dichloromethane and dibromomethane, were found in all chlorinated drinking water. A large amount of studies were carried out on the development of cancer caused by drinking water. Most studies use population data to find a relation between geographical drinking water distribution and the risk on death as a result of cancer. Other studies investigated water sources that were used by people who developed cancer and water sources that were used by people who died as a result of another disease. Because no direct measurements were used, variables were taken for surface water versus groundwater, chlorinated versus non-chlorinated water and river water carrying industrial pollutions versus river water without pollutions. These studies show there is a relation between drinking water quality and bladder-, intestinal- and anal cancer. (Cantor, 1980)

Disinfection byproducts are non-carcinogenic according to World Health Organization (WHO)

In 1991 the WHO's International Agency for Research on Cancer (IARC) evaluated the carcinogenic health risk of chlorinated drinking water based on toxicological laboratory studies and human epidemical researches. This study showed that it is hard to find a relation between the development of cancer and drinking of chlorinated water. The risk is small and cannot be proved with epidemical evidence. With all researches the estimations of exposure to disinfection byproducts were inaccurate. Furthermore, all kinds of factors are important for developing cancer, for example smoking, food, alcohol, socio-economic status and hereditary predisposition. (Disinfectants and Disinfection Byproducts, WHO,2001)

Can disinfection byproducts cause bladder cancer?

A meta-analysis of several researches shows that there is a positive correlation between exposure to disinfection byproducts in drinking water and human bladder and anal cancer. Nine percent of all cases of bladder cancer and fifteen percent of anal cancer are attributed to chlorinated drinking water and disinfection byproducts. This comes down to 10,000 cases annually. (Morris, 1992)

Risk of bladder cancer increases after lengthy exposure to chlorinated drinking water

In 1990 and 1991 in Colorado (United States) a population research was carried out on the relation between disinfection of drinking water with chlorine or chloramines and the occurrence of bladder cancer. 327 people with bladder cancer were compared to 261 people suffering from another type of cancer. On the basis of interviews and data of the Health Organization a drinking water exposure profile was created. This study showed that a relation exists between years of exposure to chlorinated drinking water and the development of bladder cancer. This risk increased after more years of exposure. After exposure of thirty years the risk on bladder cancer was 1.8 times bigger than when no exposure had occurred. The concentration trihalomethanes, nitrate and residual chlorine were not associated with the risk on bladder cancer. (McGeehin, 1993)

Fourteen to sixteen percent of bladder cancer cases is caused by disinfection byproducts

Research on lengthy exposure to disinfection byproducts in drinking water and the occurrence of bladder cancer carried out in Ontario (Canada) shows that there is a relation between lengthy exposure to disinfection byproducts and the risk on bladder cancer. The risk increased after lengthy exposure and trihalomethane concentrations of 50 μg/L or more. Fourteen to sixteen percent of all bladder cancer cases can be attributed to exposure to disinfection byproducts. (King, 1996)

Connection between exposure to disinfection byproducts and bladder cancer

In Finland research is carried out on the connection of lengthy exposure to mutagenic and carcinogenic substances in drinking water and cancer. For this study the exposure of 732 bladder cancer patients, 703 renal cancer patients and  914 other people to drinking water was determined on data on residence, water sources and historical data on water quality and water treatment. For men there was a relation between exposure and the risk on renal cancer. For women this relation was non-significant. For both men and women the connection between exposure and bladder cancer was significant. (Koivusalo, 1998)

Risk of bladder cancer is important because of the large amount of people exposed to chlorinated drinking water

A comparison of different studies to individual consumption of chlorinated drinking water and the association of bladder cancer shows there is a connection between lengthy exposure to chlorinated drinking water and bladder cancer. This risk increases after exposure for many years. This risk is not very big, but because many people are exposed to chlorinated drinking water for many years, this risk is significant because cases of bladder cancer can be attributed to disinfection byproducts. (Kogevinas, 2003)

Is there a risk of intestinal cancer due to the formation of disinfection byproducts?

Research on the connection of intestinal cancer and disinfection byproducts in drinking water show that there is an elevated risk on intestinal cancer when chlorinated drinking water is used. Marret and King examined 5000 people in Ontario (Canada), of which 950 were bladder-, intestinal or anal cancer. Data on the concentration of trihalomethanes in water were used. Other factors, including eating habits were investigated as well. This study proved that people who were exposed to concentrations of 50 μg/L or more had 1.5 times bigger risk to develop intestinal cancer. (Marret en King, 1995). 

Too little evidence on elevated risk of intestinal cancer

In 1998 a study was carried out on 685 intestinal cancer patients in Iowa (Canada). 2400 people suffering from another form of cancer were used as a control group. The concentration of trihalomethanes in drinking water were estimated. These estimations were adjusted on basis of other factors. This study showed no elevated risk on intestinal cancer. The different result of these studies can be a coincidence or be caused by another composition of the drinking water or other factors. In this study there is too little evidence for a relation between exposure to disinfection byproducts and an elevated risk on intestinal cancer. (Mills, 1998)

Do disinfection byproducts elevate the risk of anal cancer after long-period exposure to chlorinated drinking water?

A study carried out in Iowa (USA) in 1986 and 1989 with data from intestinal and anal cancer patients shows there is no elevated risk on intestinal cancer after long time exposure to chlorinated drinking water or trihalomethanes. For anal cancer there is an elevated risk however. This risk is even bigger for people who eat little fibrous food. A lack of physical exercise also elevates the risk on anal cancer. (Hildesheim, 1998)

Do disinfection byproducts influence the reproduction and development of humans?

Most attention on health effects of disinfection byproducts is on cancer caused by lengthy exposure to disinfection byproducts in drinking water. Standards that are being used for permitted concentrations of disinfection byproducts are based on carcinogenic abilities of these substances. (Singer, 1999)

Is there a connection between affects on laboratory animals and humans?

Laboratory tests with animals show that exposure to disinfection byproducts during pregnancy influence reproduction and development and induce birth-defects and spontaneous abortion. With humans these effects have been investigated with population data on drinking water quality, water treatment and birth data. The concentrations disinfection byproducts that cause these effects are in most cases many times bigger than concentrations that can cause cancer after lengthy exposure. To regulate disinfection byproducts in drinking water all potential health effects have to be considered. (Singer, 1999)

Did research of the effects on reproduction and development of humans take place?

The number of epidemiological studies on exposure to disinfection byproducts and the influence on reproduction and birth defects is small. However, these studies show there is a connection between exposure to trihalomethanes and spontaneous abortion, birth defects and growth delay. (Wigle, 1998)

Is there a connection between chlorine dioxide byproducts, low birth-weight and premature birth?

To prevent the formation of chlorinated carcinogenic disinfection byproducts, other disinfectants are used. These disinfectants also produce disinfection byproducts that can be harmful for human health. Chlorine dioxide for example produces disinfection byproducts chlorite and chlorate, which have health effects on vulnerable people as newborn babies. Disease and death rates of newborns in two communities were studied. In one community water was disinfected with chlorine, the other used water disinfected with chlorine dioxide. The relation between exposure to water treated with chlorine dioxide of the mother during pregnancy, premature birth and a low birth weight was significant. There was no difference in the number of birth defects and still births. (Tuthill, R. 1982)

Must the exposure rate has be determined more adequately?

On the basis of available epidemiological data research has been carried out on the relation between disinfection byproducts in drinking water and effects on reproduction and development. The epidemiological evidence on a relation between exposure to disinfection byproducts and development is weak. If a connection is found, one has to be careful in drawing any conclusions. The research methods being used are very diverse and it is difficult to compare results. Future studies have to use enhanced methods to determine exposure. This can be achieved by using exposure markers and taking seasonal and annual differences in concentrations of disinfection byproducts through different transmission routes into account. Furthermore population research is required to determine male and female fertility, growth delay and specific birth defects. (Reif, 1996)

Is there a connection between exposure to chlorinated drinking water and low birth weight?

A research was carried out on exposure during pregnancy to chlorinated drinking water with a high amount of natural organic matter and non-chlorinated drinking water with a small amount of natural organic matter. Birth data from 137,145 Norwegian births between 1993 and 1995 were used. The study showed no connection between exposure to chlorinated drinking water and a risk for low birth weight and small body length. The risk on premature birth was slightly smaller with exposure to chlorinated drinking water than non-chlorinated drinking water. (Jaakkola, 2001)

Is there a risk of birth defects after exposure to disinfection byproducts in drinking water?

In Norway a research was carried out on the relation between specific birth defects and the occurrence of natural organic matter and disinfection byproducts in drinking water. Birth data from 285,631 births in Norway from 1993-1998 were used. The risk on birth-defects and more specific heart-, breathing and urine tract defects were associated with exposure to disinfection byproduct during pregnancy. The risk on abdominal wall defects increases significantly after higher exposure. (Bing-Fang, 2002)

What are the effects of exposure to trihalomethanes in drinking water on foetal development?

Data from 56,513 births in Massachusetts (USA) in 1990 were used to investigate the effect of exposure to trihalomethanes in drinking water on foetal development. Exposure causes a low birth weight and small body length, also known as foetal growth delay. Comparison of trihalomethane concentration show that 80µg/l or more lower the birth weight with 32 gram. No evidence was found on exposure to trihalomethanes and premature birth. (Wright, 2003)

Is there evidence for an influence of disinfection byproducts on reproduction?

For this research the epidemiological and toxicological evidence of studies on effects of disinfection byproducts on reproduction were weighed. There was too little evidence on a relation between exposure to disinfection byproducts in drinking water during pregnancy and effects on foetal development. Effects that were investigated were birth weight, premature delivery, some congenital defects and early death of the newborn. There was little evidence for defects on the central nervous system, spinal cord, spontaneous abortion and stillbirth. There was sufficient evidence for a relation between growth delay, and defects on urine tracts and exposure to disinfection byproducts. The epidemiological research that has been carried out so far is in-efficient in proving a connection between disinfection byproducts and reproduction effects. To see if there is an evidence the true amount of water and disinfection byproducts that women consume has to be measured. (Graves, 2001) Other scientists reached this same conclusion. (Nieuwenhuijsen, 2000)

Is the risk of environmental pollutions on birth defects measurable?

Research on the development of birth defects as a result of exposure to chemical environmental pollutions (drinking water pollutions, pesticides, waste, industrial pollutions, food pollutions and disasters with a large emission of chemical pollutions) show that it very difficult to determine the potential risk for birth defects caused by exposure to environmental pollutions. To prevent birth defects exposure to all chemical environmental pollutions should be prevented. (Dolk, 2003)

Is there a relation between exposure to bromodichlormethane, chloroform and birth defects?

On the basis of birth data of births in Nova Scotia (Canada) from 1988 to 1995 and results of water monitoring tests research has been carried out on birth effects of bromodichlormethane and chloroform. Exposure during pregnancy to bromodichloromethane concentrations of 20 or more µg/l were associated with an elevated risk on defects on the neural tube. Exposure to chloroform points out to an elevated risk of chromosomal defects. The results of this study show that research on the relation between specific disinfection byproducts and birth defects is needed. (Dodds, 2001)

Is there a relation between exposure to trihalomethanes and chlorine dioxide and birth defects?

In 2001 in Sweden a research was carried out on the relation between heart- and artery defects in children and trihalomethane concentrations in drinking water before and during pregnancy. 753 out of 59,422 children suffered from heart and artery defects. The risk on these effects was elevated when both chlorine dioxide and hypochlorite were used. This risk was more high than when solely hypochlorite was used. Chlorine dioxide seems to increase the risk on heart and artery defects. This can also be caused by the composition of the water, which might have been more polluted. All water investigated had trihalomethane levels lower than the standard levels. This points out that even below these concentrations effects on reproduction take place. (Cedergren, 2001)

Do chlorine dioxide or hypochlorite influence physical parameters?

In Italy research was carried out on physical parameters at birth and the relation with drinking water disinfected with chlorine dioxide or hypochlorite. This research was done because of earlier publications on birth effects of disinfectants and disinfection byproducts. Data of 548 women from Genua (Italy) giving birth from 1988 or 1989 and using water disinfected with sodium hypochlorite or chlorine dioxide were used. Data from 128 women from nearby Chiavari, were used as a control group. Their drinking water was not disinfected. Other factors influencing birth-defects were also investigated. There was a relation between drinking water disinfected with chlorine dioxide and small body length and cranial span. A hypothesis for this result is that the immunity of women exposed to chlorine dioxide is decreased. (Kanitz, 1996)
The outcome of this study can be questioned, because most Italians use bottled water, which is disinfected with ozone. The chlorine dioxide, chlorite and chlorate concentration has not been measured, and no dose-effect relationship can be established. 

Are the studies of reproduction effects reliable?

The results of these studies show that there is probably a relation between exposure to (chlorinated) disinfection byproducts before and during pregnancy and birth-defects. A low birth weight and growth delay are mostly found. The evidence for spontaneous abortion, birth-defects and still-birth is not very consequent. The evidence is not strong enough to show a dose effect relation. This can be caused by research methods and techniques being used. 

What do new studies imply?

The amount of water that women use has to measured. Then the concentration disinfection byproducts can be determined more accurately. It would even be better to determine the concentration disinfection byproducts from the tap. The composition and concentration of disinfection byproducts can be changed under influence of pH, temperature and contact time in water distribution network compared to the concentration and composition at the water company. 
It is not clear yet whether all disinfection byproducts cause health effects and whether their effects differ. Further research is required.
Other factors, like smoking and exposure to environmental pollutions have to be investigated as well. 

What recommendations can be made for future research on health risks of disinfection byproducts in drinking water?

The microbiological quality of drinking water should be maintained while preventing the formation of disinfection byproducts. Efficient disinfection is preferred. Health risks of disinfection byproducts are small compared to health risks of waterborne diseases. This is proven by the cholera-epidemic that occurred in Peru in 1991. This epidemic was caused by inadequate drinking water disinfection. Worldwide attention for disinfection byproducts and the great number of scientific articles on disinfection byproducts caused many drinking water suppliers in South-America to stop drinking water chlorination. The acute health risk of pathogenic micro-organisms in drinking water is much higher, about 100.000 to 1.000.000 times higher than the risk of long-term exposure to disinfection byproducts. The cholera-epidemic spread to all 19 South-American countries and caused 1.200.000 patients and 40.000 deaths. (WHO, 1994)
Health risks of disinfection byproducts are very low at concentrations found in drinking water. Nevertheless these risks can not be ignored, because of the large number of people exposed to disinfection byproducts. There are still a lot of disinfection byproducts that must be identified. Health risks must be researched, as well. The effects of mixtures of disinfection byproducts must be researched. Some disinfection byproducts may be mutagenic and must be researched, as well.

What methods can be used to control disinfection byproducts?

Changing the point of disinfectant application, using an alternative disinfectant, removing natural organic matter that produces disinfection byproducts in combination with disinfectants and removal of disinfection byproducts after disinfection can be used to control disinfection byproducts. 

In general it is best to remove as much organic matter as possible from water, before disinfection is applied. This can be achieved with existing water treatment techniques. Coagulation is used to remove particles and turbidity. Active coal can be used to absorb organic substances. Membranes can be applied to remove organic matter from water. 

Alternative disinfectants, for example ozone, chlorine dioxide, potassium permanganate and chloramines can also be used to prevent the formation of disinfection byproducts. However all disinfectants produce disinfection byproducts. Chlorinated disinfection byproducts are researched more thoroughly than other byproducts. (Singer, 1999)

What are the standards for disinfection byproducts?

Some disinfection byproducts are considered harmful for public health (chloroform, dibromochloromethane and bromoform are probably carcinogenic and dichlorobromomethane, dichloroacetonitrile and chloral hydrates are possibly carcinogenic). Health institutions worldwide have set standards for the maximum concentration of disinfection byproducts in drinking water. 

EU
In the European Drinking Water Directive 98/83/EC (1998) the maximum standard for trihalomethanes is set to 100 μg/L. If possible countries should strive for lower concentrations. 

WHO
The WHO describes separate standards for three trihalomethanes:
- bromodichloromethane (BDCM) 60 μg/L
- bromoform 100 μg/L
- chloroform 200 μg/L.

USA
The EPA is concerned with regulation on disinfection byproducts in the United States since 1979. In 1996 the Safe Drinking Water Act was revised and the Congress asked EPA to set new standards for disinfectants and disinfection byproducts. This revision aims at lowering the health risk of disinfection byproducts, while protecting microbiological quality of the water. In 1998, EPA promulgated the Stage 1 Disinfectants and Disinfection Byproducts Rule. The standard on total trihalomethane concentration is 80 μg/L and for halogenated acidic acid 60 μg/L. The guideline also states that advanced coagulation must be used to remove organic matter. (EPA, 2001)