DBPs are formed in drinking water treatment and distribution have been associated with human health risks, but the interpretation of disparate and inconclusive results has been complicated by the inability to establish toxicological similarity of drinking waters studied (Bull, Rice et al. 2009). The equivocal nature of the health effects evidence suggests a need for new scientific and policy approaches to DBP risk assessment. First, we present an overview, then some of the formation and DBP history. Next, we talk about conflicting health evidence, with some suggestion that bromine incorporation alters toxicological characteristics of the mixtures. Finally, we integrate occurrence information with this toxicological and epidemiological conundrum to suggest further investigation may be warranted.
Disinfection by-products (DBPs) are formed when water supplies are disinfected using chemical disinfectant (e.g., chlorine, chloramines, chlorine dioxide) in order to protect consumers from disease outbreak related to microbial contaminants. Generally, disinfection is a two-step process. Primary disinfection is used to inactivate microorganisms that may have been present in the source water and have survived the physical and chemical treatment processes. Secondary disinfection involves addition of sufficient disinfectant to the water as it enters the distribution system such that the residual disinfectant concentration will ensure microbial stability of the water delivered to the consumer tap. Although some water utilities employ alternative disinfection methods, such as ultraviolet (UV) disinfection and ozonation, DBPs can occur from all disinfection methods, except UV (Minear and Amy 1996; Xie 2003). Disinfectant chemicals may also be used in other processes such as enhanced coagulation to improve the removal of natural organic matter and other suspended solids, taste and odor control, or pre-oxidation to improve the removal of some metals or subsequent filtration process performance (Xie 2003).
DBPs can occur from all disinfection methods, except UV
Since researchers first identified DBPs in US drinking water supplies, concerns have emerged about their potential health effects. Originally focusing on their carcinogenic potential, initial investigations highlighted the challenge of achieving protection against pathogenic risks while reducing exposures to the newly suspected carcinogenic disinfection-by-products. The dominant risk assessment culture at the time was conducive for the precautionary regulation of trihalomethanes (THMs), the most prevalent class of DBPs by mass. It would not be until the late 1980s and early 1990s that the association between other potential health outcomes, most notably adverse reproductive outcomes, would be considered (Reif, Bachand et al. 2000 ). The first precautionary regulations promulgated in 1979, and two recent stages of regulation in 1998 and 2006 (EPA 1998; EPA 2006), are policy products of sustained public health risk construction and research. The 1998 Stage 1 Disinfectants/Disinfection-Byproducts (D/DBP) rule and the Interim Enhanced Surface Water Treatment Rule (EPA 1998) attempted to refine the regulatory balance between microbial protection and health risks associated with DBP exposure, establishing maximum contaminant levels (MCLs) for haloacetic acids (HAAs), trihalomethanes (THMs), bromate, and chlorite. The Stage 1 D/DBP rule has been extended by EPA’s Stage 2 D/DBP rule in 2006 (EPA 2006). This rule targets reducing short-term exposures to address potential reproductive and developmental health risks in addition to cancer risks posed by D/DBP exposure.
Because epidemiological and toxicological evidence supporting the Stage 1 and Stage 2 rules was equivocal (Graves, Matanoski et al. 2001; Tardiff, Carson et al. 2006), it was thought that improved epidemiological evidence might yield better informed decision making when these standards are reviewed. Currently, some researchers believe that much of the DBP mixture risk is due to non-regulated and uncommonly measured compounds (Hamidin, Yu et al. 2008; Rice, Teuschler et al. 2009). Consequently, new approaches are being developed for consideration in the assessment of risks posed by DBP exposures (Rice, Teuschler et al. 2009). Currently, DBPs are regulated as mixtures of four individual trihalomethanes (THM4; chloroform, bromodichloromethane, dibromochloromethane, bromoform) and five haloacetic acids (HAA5; trichloroacetic acid, dichloroacetic acid, monochloroacetic acid, dibromoacetic acid, and monobromoacetic acid) under the assumption of toxicological additivity with respect to bladder cancer and potential adverse reproductive outcomes, a non-cancer health endpoint (Hurtzberg, Choudhury et al. 2000; Simmons, Richardson et al. 2002). One may intuitively assume that such a mixtures approach may provide an appropriate model of human exposure, and thus a realistic model of associated risks. This notion is difficult to test. For example, it is not technically feasible to evaluate all the combinations of even the known DBPs to evaluate the assumption of additivity for reproductive outcomes (Simmons, Richardson et al. 2002).
"Although the concept of using the BIF as a mixture-based DBP regulatory metric might potentially be useful, substantial obstacles to its practical application must be addressed."
More recently, occurrence studies have found that many non-regulated DBPs may occur at concentration levels similar to their regulated analogues, especially for the HAA class (Krasner, Westerhoff et al. 2009). These findings hold especially for drinking waters whose source is impacted by bromide. When bromide reacts with chlorine or chloramines, bromide is oxidized to hypobromous acid. Hypobromous acid then reacts with natural organic matter analogously to hypochlorous acid, and is incorporated into DBPs. Because hypobromous acid reacts faster than hypochlorous acid, bromine substitutes for chlorine in DBP molecules, altering their molecular weight and chemical properties. Two consequences are most important for drinking water treatment: first, bromine-substituted DBPs are heavier, potentially confounding the compliance status of utilities treating bromine impacted source waters; second, bromine-substituted DBPs may be more toxic than their chlorine-only analogues (Muellner, Wagner et al. 2007).
These occurrence findings are also of great importance for interpreting and synthesizing health effects evidence to support policy. The main implication of the occurrence findings is to potentially reduce the inference of external validity of health effects studies by making it difficult to evaluate the toxicological similarity of drinking water DBP mixtures obtained from different sources and treatment techniques. In fact, this difficulty is being undertaken by an ambitious EPA research agenda in which researchers attempt to compile a suite of chemical characterization factors that facilitate the assessment of toxicological similarity (Claxton, Pegram et al. 2008; Crosby, Simmons et al. 2008; Miltner, Speth et al. 2008; Narotsky, Best et al. 2008; Rice, Teuschler et al. 2008; Richardson, Thruston Jr. et al. 2008; Simmons, Richardson et al. 2008). This research could have significant implications for re-evaluation of the Stage II D/DBP rule, because the THM4 and HAA5 based regulation suggests that DBP sum concentrations of the species listed above are an adequate measure of DBP mixture toxicological similarity.
However, one of the chemical characterization factors suggested by the EPA researchers is the bromine incorporation fraction (BIF) (Bull, Rice et al. 2009). In addition, the BIF could possibly be a useful regulatory metric, complementing the use of THM4 and HAA5 concentrations. First, the BIF is influenced by treatment process configurations and source water quality, including the bromide to total organic carbon ratio at the first point of chlorine addition, the chlorine applied through the plant to influent total organic carbon ratio, alkalinity at the first point of chlorine addition, specific UV absorption at the first point of chlorine addition, and free chlorine residual (Francis, VanBriesen et al. 2009). Second, Obolensky and Singer (2005) and Francis et al. (2009) show that the BIF also encodes information transcending DBP subclasses because class-based BIFs are correlated among each other. As a result, the BIF may communicate information about the DBP compounds that are not measured routinely. Third, the BIF may provide more information than current compliance measures alone at modest additional cost. While the bromine incorporation fraction is positively correlated with concentrations of species containing at least two bromine atoms, this correlation decreases as the number of chlorine atoms in a DBP species increases (Francis, Small et al. 2009). The bromine incorporation fraction is negatively correlated with fully chlorinated species and DBP class sum concentrations, including the regulated class sum concentrations of THMs and HAA5. As a result, it has been hypothesized that the bromine incorporation fraction from a fully measured class (i.e., THM bromine incorporation fraction) may be used to predict bromine-substituted species from other classes for which at least one fully chlorinated species has been measured (Bull, Rice et al. 2009). If correct, this would allow use of limited analytical results to extrapolate to health effects from complex mixtures of DBPs. This, in turn, may permit more cost-effective use of limited DBP occurrence information, as routine measurement of DBPs from each prioritized class is currently cost-prohibitive.
"Despite these considerable obstacles, the study of DBP mixture metrics has emerged. "
Although the concept of using the BIF as a mixture-based DBP regulatory metric might potentially be useful, substantial obstacles to its practical application must be addressed. The primary obstacle is the structured risk evaluation and cost benefit analysis necessary to be introduced as a regulatory metric. Driedger calls this process the “construction and contestation of scientific evidence concerning chlorination DBP” risks (Driedger, Eyles et al. 2002), demonstrating that the interactions between epidemiologists and toxicologists dictate the boundary setting and legitimization of science authorities in the DBP health effects debate. It is very difficult, to say the least, to demonstrate adverse toxicological effects attributable to the BIF of a DBP mixture in controlled laboratory studies. It is nearly impossible to construct a hypothesis for investigating the risks in human populations using the BIF if plausible causal relationships between BIF and adverse effects cannot be identified in toxicological studies.
Despite these considerable obstacles, the study of DBP mixture metrics has emerged. Researchers have found potential value in the creativity engendered by the BIF in addressing the problem of complex DBP mixtures. In future research, the use of the bromine incorporation fraction might be introduced under guidance from concepts influenced by the Four Lab Study, especially subsequently developed statistical techniques (Feder, Ma et al. 2009; Feder, Ma et al. 2009). Indeed, the bromine incorporation fraction could be an important metric for establishing toxicological similarity of DBP mixtures in the absence of evidence to justify routine measurement of additional DBP classes in epidemiological studies and occurrence surveys. Consequently, development of a framework for setting research priorities incorporating DBP mixture metrics might be a productive activity that may facilitate the assimilation of the research described in this opinion article.
Bull, R. J., G. Rice, et al. (2009). "Determinants of whether or not mixtures of disinfection by-products are similar." Journal of toxicology and environmental health. Part A 72: 437-460.
Bull, R. J., G. E. Rice, et al. (2009). "Chemical measures of similarity among disinfection byproduct mixtures." Journal of Toxicology and Environmental Health, Part A 72(7): 482-493.
Claxton, L. D., R. Pegram, et al. (2008). "Integrated disinfection by-products research: Salmonella mutagenicity of water concentrates disinfected by chlorination and ozonation/postchlorination." Journal of Toxicology and Environmental Health, Part A 71: 1187-1194.
Crosby, L. M., J. E. Simmons, et al. (2008). "Integrated disinfection by-products (DBP) mixtures research: gene expression alterations in primary rat hepatocyte cultures exposed to DBP mixtures formed by chlorination and ozonation/postchlorination." Journal of Toxicology and Environmental Health, Part A 71: 1195-1215.
Driedger, S. M., J. Eyles, et al. (2002). "Constructing scientific authorities: Issue framing of chlorinated disinfection byproducts in public health." Risk Analysis 22(4): 789-802.
EPA, U. S. (1998). National Primary Drinking Water Regulations: Disinfectants and Disinfection Byproducts. Federal Register, Federal Register. 2040-AB82: 69389-69476.
EPA, U. S. (2006). National primary drinking water regulations: Stage 2 disinfectants and disinfection byproducts rule. U. S. E. P. Agency, Federal Register. 2040-AD38: 387-493.
Feder, P. I., Z. J. Ma, et al. (2009). "Evaluating sufficient similarity for drinking-water disinfection by-product (DBP) mixtures with bootstrap hypothesis test procedures." Journal of Toxicology and Environmental Health, Part A 72(7): 494-504.
Feder, P. I., Z. J. Ma, et al. (2009). "Evaluating sufficient similarity for disinfection by-product (DBP) mixtures: multivariate statistical procedures." Journal of Toxicology and Environmental Health, Part A 72: 468-481.
Francis, R. A., M. J. Small, et al. (2009). "Multivariate distributions of disinfection byproducts in chlorinated drinking water." Water Research 43: 3453-3468.
Francis, R. A., J. M. VanBriesen, et al. (2009). Bayesian statistical modeling of disinfection byproduct (DBP) bromine incorporation in the ICR database. Pittsburgh, PA, Carnegie Mellon University.
Graves, C. G., G. M. Matanoski, et al. (2001). "Weight of evidence for an association between adverse reproductive and developmental effects and exposure to disinfection by-products: a critical review." Regul Toxicol Pharmacol 34(2): 103-124.
Hamidin, N., Q. J. Yu, et al. (2008). "Human health risk assessment of chlorinated disinfection byproducts in drinking water using a probabilistic approach." Water Research 42: 3263-3274.
Hurtzberg, R., H. Choudhury, et al. (2000). Supplementary Guidance for Conducting Health Risk Assessment of Chemical Mixtures. Washington, DC, Risk Assessment Forum Technical Panel
US Environmental Protection Agency.
Krasner, S. W., P. Westerhoff, et al. (2009). "Occurrence of Disinfection Byproducts in United States Wastewater Treatment Plant Effluents." Environmental Science & Technology 43: 8320-8325.
Miltner, R. J., T. F. Speth, et al. (2008). "Integrated disinfection by-products mixtures research: disinfection of drinking waters by chlorination and ozonation/postchlorination treatment scenarios." Journal of Toxicology and Environmental Health, Part A 71: 1133-1148.
Minear, R. A. and G. L. Amy (1996). Disinfection By-Products in Water Treatment: the chemistry of their formation and control, CRC Press.
Muellner, M. G., E. D. Wagner, et al. (2007). "Haloacetonitriles vs. Regulated Haloacetic Acids: Are Nitrogen-Containing DBPs More Toxic?" Environ. Sci. Technol. 41(2): 645-651.
Narotsky, M. G., D. S. Best, et al. (2008). "Integrated disinfection by-products mixtures research: assessment of developmental toxicity in Sprague-Dawley rats exposed to concentrates of water disinfected by chlorination and ozonation/postchlorination." Journal of Toxicology and Environmental Health, Part A 71: 1216-1221.
Obolensky, A. and P. C. Singer (2005). "Halogen Substitution Patterns among Disinfection Byproducts in the Information Collection Rule Database." Environ. Sci. Technol. 39(8): 2719-2730.
Reif, J. S., A. Bachand, et al. (2000). Reproductive and developmental effects of disinfection by-products. Fort Collins, CO, Department of Environmental Health, Colorado State University.
Rice, G. E., L. K. Teuschler, et al. (2009). "Evaluating the similarity of complex drinking waer disinfection byproduct mixtures: Overview of the issues." Journal of Toxicology and Environmental Health, Part A 72(7): 429-436.
Rice, G. E., L. K. Teuschler, et al. (2008). "Integrated disinfection by-products research: assessing reproductive and developmental risks posed by complex disinfection by-product mixtures." Journal of Toxicology and Environmental Health, Part A 71: 1222-1234.
Richardson, S. D., A. D. Thruston Jr., et al. (2008). "Integrated disinfection by-products mixtures research: comprehensive characterization of water concentrates prepared from chlorinated and ozonated/postchlorinated drinking water." Journal of Toxicology and Environmental Health, Part A 71: 1165-1186.
Simmons, J. E., S. D. Richardson, et al. (2002). "Development of a Research Strategy for Integrated Technology-Based Toxicological and Chemical Evaluation of Complex Mixtures of Drinking Water Disinfection Byproducts." Environmental Health Perspectives 110(Supplement 6): 1013-1024.
Simmons, J. E., S. D. Richardson, et al. (2008). "Research issues underlying the four-lab study: integrated disinfection by-products mixtures research." Journal of Toxicology and Environmental Health, Part A 71: 1125-1132.
Tardiff, R. G., M. L. Carson, et al. (2006). "Updated weight of evidence for an association between adverse reproductive and developmental effects and exposure to disinfection by-products." Regul Toxicol Pharmacol 45(2): 185-205.
Xie, Y. F. (2003). Disinfection Byproducts in Drinking Water: Formation, Analysis, and Control. Washington, D.C., CRC Press.