Invited Perspective: Drinking Water Disinfection By-Products and Cancer—A Historical Perspective

The year 2024 has been an anniversary for research on disinfection by-products (DBPs). It marks 50 years since two independent studies first reported the formation of trihalomethanes (THMs) through chlorine reactions with naturally occurring organic matter in water.^1,2 These were seminal studies, stimulating prolific research on the analytical chemistry, toxicology, exposure science, and epidemiology of DBPs. Half a century later, there are still so many unanswered questions.

New research by Helte et al.,³ published in this issue of Environmental Health Perspectives, represents a milestone in DBP and cancer research. It exhaustively synthesizes the state of the knowledge and provides robust estimates for exposure–response functions for bladder and colorectal cancer, the two cancer sites with the most consistent evidence. Helte et al. provide the first exposure–response function for colorectal cancer and THMs based on a systematic review and an exposure–response function for bladder cancer updating previous efforts.^4,5 This is a very valuable contribution that—assuming a causal association—will allow health impact assessment exercises to estimate the burden of disease attributable to this prevalent exposure. The systematic review and derived exposure–response functions are desperately needed if scientific evidence is to be translated into plainer language for policymaking, regulations, economic evaluations, and so on.

DBPs are a complex exposure. Approximately 600 DBP molecules have been identified,⁶ although the subset that has been quantified constitutes only $\sim 30\%$ of total organic halogens (TOX).⁷ A chemically diverse group, some DBPs (including THMs) are volatile and skin permeable, involving different exposure routes. Drinking water supplies need to be disinfected to be microbiologically safe; moreover, we are exposed to tap water while showering, bathing, washing dishes, or swimming in a pool.⁸ Thus, DBPs are practically a universal exposure in the general population on public drinking systems, and exposure assessment in epidemiologic research is challenging, to say the least.

The biological mechanisms linking DBP exposure and cancer are not completely understood, and the responsible DBPs, either individuals or mixtures, remain unidentified. However, there is enough toxicologic and epidemiologic evidence indicating that cancer is a biologically plausible outcome of long-term DBP exposure. Epidemiologic research on DBPs and cancer has substantially evolved from a first generation of ecological studies in the 1970s (e.g., Page et al.,⁹ Kuzma et al.,¹⁰ Cantor et al.¹¹) to a second generation of cancer mortality case–control studies using registries and crude exposure surrogates in the 1980s (e.g., Gottlieb and Carr,¹² Kanarek and Young,¹³ Zierler et al.¹⁴). Based on these studies, the World Health Organization International Agency for Research on Cancer (IARC) evaluated human carcinogenicity of chlorinated drinking water in 1991.¹⁵ Inadequate human evidence of carcinogenicity led to the conclusion that chlorinated drinking water is a group 3 agent—that is, not classifiable as to its carcinogenicity to humans. This was not surprising given the nonspecific nature of “chlorinated drinking water.”

In the 1990s and 2000s, epidemiologic studies (e.g., King and Marret,¹⁶ Cantor et al.,¹⁷ Villanueva et al.,⁸ Beane Freeman et al.¹⁸) incorporated the use of THMs as a marker of DBP exposure, collected residential histories from study participants, and evaluated incident cancer cases. The use of THMs for exposure assessment was justified for two reasons. First, these chemicals are the most abundant by-product of chlorination⁷ (in hypochlorite form, the most widespread disinfectant at that time^7,19). Second, THMs were the first regulated and routinely monitored DBPs.¹⁹ Historical records available in high-income countries allowed long-term exposure assessment in epidemiologic studies. However, THMs are not the most toxic DBPs,⁷ and they have limitations as a marker of other DBPs,²⁰ so questions remained on the biological mechanisms linking DBP exposure with cancer risk.

The new generation of studies incorporates other DBPs, such as haloacetic acids, or simultaneously evaluates THMs along with other carcinogens in drinking water, such as nitrate.^21–24 This represents a clear improvement over previous studies, but still there is a lack of evidence on more toxic, less abundant, and unregulated DBPs, such as haloacetonitriles, $n$-nitrosodimethylamine, or chlorate and chlorite (which were recently regulated in Europe).²⁵ These, among many other DBPs, are formed by alternative disinfectants (e.g., chloramine, chlorine dioxide) that are used to reduce THM formation and thus comply with regulations.

The mismatch between epidemiologic evidence on THMs as a class of DBPs and IARC’s agent-based assessments explains why most of the evaluated DBPs have been classified as group 3 or 2B (possibly carcinogenic to humans). IARC conclusions have been based on animal data because human evidence is limited or lacking. Group 3 DBPs include dibromochloromethane, bromoform, chloroacetonitrile, dichloroacetonitrile, trichloroacetonitrile, bromochloroacetonitrile, and chlorite.^26,27 Group 2B DBPs include chloroform, bromodichloromethane, dichloroacetic acid, trichloroacetic acid, bromochloroacetic acid, dibromoacetic acid, dibromoacetonitrile, mutagen X, and bromate.^27–30 Only chloral hydrate is classified as 2A (probably carcinogenic to humans), based on mechanistic evidence.²⁹ This scenario may change in the coming years because IARC has included DBPs in the priority list of agents to be evaluated in the next 5 y.³¹ However, the single-chemical approach does not represent real-world exposure, and human epidemiologic evidence will probably remain limited for unregulated DBPs.

As the DBP scientific community celebrates the publication by Helte et al.,³ I feel my own sense of connection to this milestone in DBP research; it coincides with a special celebration just around the corner—my own 50th birthday in 2025. DBP research represents an important part of my professional life, with the past 26 years immersed in this field, seeming as though my life and the trajectory of DBP research have been moving in parallel, each marking significant milestones along the way.

Conclusions and opinions are those of the individual authors and do not necessarily reflect the policies or views of EHP Publishing or the National Institute of Environmental Health Sciences.

Refers to https://doi.org/10.1289/EHP14505