By Richard C. Pleus, Ph.D.

Summary: Oftentimes, knowing the process for how a drinking water standard—or any policy—will be calculated is crucial to being able to have your voice heard in a timely manner to effectively manage the issue in your best interest. Such has been the case recently with proposals for new radon and arsenic rules and several other recent or impending new rules underscore that necessity. Following is a discussion on how health risk factors are assessed and implications for the water treatment industry.

What can perchlorate ion (ClO4-)—an oxidizer in solid propellant systems for rockets such as the space shuttle—tell us about chlorine dioxide (ClO2) and its disinfection by-products (DBPs), specifically chlorate (ClO3-) and chlorite (ClO2-) ions? Plenty! This article will point out similarities of chlorine dioxide and its DBPs to what has been happening with perchlorate. More specifically, these chemicals are found in drinking water supplies, affect the thyroid gland, are the primary drivers in establishing toxicity guideline values by the federal government, are oriented to children’s health and the chemical’s mode of action.

One reason for this issue’s timeliness relates to research interests in DBPs of chlorine dioxide by the U.S. Environmental Protection Agency (USEPA) and recent attention perchlorate has received as the agency moves toward revising the reference dose (RfD) of perchlorate. Also, the high production volume (HPV) chemical initiative—see—will likely result in large amounts of toxicity testing requiring new data on these compounds. Lastly, our experience assisting industry in revision of the perchlorate RfD might provide valuable insight to manufacturers and users of chlorine dioxide if this USEPA trend of requiring more and more toxicology data continues.

DBPs & improved detection
Perchlorate, chlorine dioxide, ClO3– and ClO2– are found in drinking water sources around the country and the world. Application of a sensitive new detection method revealed groundwater contamination in a number of states, particularly Utah, California, Nevada and Arizona, suggesting widespread human exposure to these compounds. Chlorine dioxide has become increasingly popular as a microbial disinfectant because it reportedly provides better treatment of water with odor and taste problems and more effectively reduces elevated levels of iron, manganese, hydrogen sulfide and phenolic compounds than chlorine gas. According to the USEPA,2 chlorine dioxide is used in about 13 percent of U.S. drinking-water treatment facilities. Depending on the pH of the water, an estimated 30 percent and 70 percent will be reduced to ClO3– and ClO2-, respectively (see Reaction 1).

In the mid-1970s, researchers discovered the reaction of the chlorine ion with organic and inorganic compounds in water produced DBPs. Examples of organic DBPs include halogenated organics such as trihalomethanes (THMs), aldehydes and ketones. Inorganic DBPs include ClO2– and ClO3-. Pursuant to requirements of the Safe Drinking Water Act (SDWA) Amendments of 1996, the USEPA is using regulatory activity—in the form of the Microbial/Disinfection By-Products Rule—to control DBPs in water.

Toxicity guidelines
For perchlorate and many other chemicals, the RfD is the fundamental toxicity guideline value upon which safe drinking water levels (SDWLs) or soil clean-up values are set. An RfD is “an estimate…of a daily oral exposure to the human population (including sensitive subgroups) that is likely to be without an appreciable risk of deleterious effects during a lifetime.”3 In addition, some of these compounds have other surrogate toxicity guideline values besides RfDs. For example, the USEPA set a maximum contaminant level goal (MCLG) and a maximum contaminant level (MCL) for ClO2-, as well as a maximum residual disinfectant level goal (MRDLG) and a maximum residual disinfectant level (MRDL) for chlorine dioxide. MCLGs and MRDLGs are non-enforceable health goals for public water systems where no known or anticipated adverse effect on the health of persons is expected to occur and that allow an adequate margin of safety. MCLs and MRDLs are enforceable standards set as close to the MCLGs and MRDLGs as technically and economically possible (see Table 1).

Children & observable effects
While chlorine dioxide, ClO2– and ClO3– ions can potentially affect a number of organ systems, the focus here is on thyroidal effects in development of the fetus and young children. The toxicological and medical literature for chlorine dioxide and ClO2– and ClO3– ions regarding their non-cancer thyroidal effects is small and shows inconsistencies. Many studies show thyroid-related effects from exposure to chlorine dioxide and ClO2– and ClO>sub>3- ions conducted in laboratory animals in the 1980s. Several criticisms of these studies have surfaced including questions about experimental design and exposure duration. For example, the experimental design of most of these studies isn’t particularly helpful for determining public health guideline values, such as an RfD. Because of this, the number and quality of studies in the chlorine dioxide and ClO2– and ClO3– database makes it difficult to determine doses, i.e., a no-adverse-effect-level (NOAEL) for critical health endpoints such as for thyroidal effects.

Another criticism of these animal studies is the exposure duration. For the majority of these studies, the duration was short, lasting only weeks. Typically, animal studies for chronic exposures are run for two years using a range of doses. Similar criticisms were made about the number and quality of studies of perchlorate in 1997. At that time, scientists recommended animal studies that would fill the data gaps in the perchlorate database. Since then, we have seen a skyrocketing rise in the number of studies where perchlorate was administered to animals and humans. All of these studies will now be used to revise the RfD.

Chronic thyroid exposure
Although the database of low-dose, chronic-exposure studies for chlorine dioxide, ClO3- and ClO2– ions is small, some of these studies report thyroidal effects. The thyroid gland traps the iodide ion from blood and, in turn produces the thyroid hormones triodothyronine (T3) and thyroxine (T4). Iodide is critical for the action of these hormones. Thyroid hormones are under the control of the thyroid-stimulating hormone (TSH), which is produced and released from the anterior pituitary gland. If the level of thyroid hormones falls in the blood, TSH is secreted, which in turn stimulates the thyroid gland to release T3 and T4. In adults, T3 and T4 control metabolism and growth. In the developing fetus and newborn, these hormones are critical for nervous system development. It’s these thyroidal effects on the development of the fetal and neonatal nervous system that piqued the interest of the USEPA, particularly in light of the agency’s recent focus on child health and the mode of action of environmental chemicals.

In some instances, there may be studies conducted with similar dose ranges and exposures that don’t show the same effects. This appears to be true for chlorine dioxide, ClO2- and ClO3-. The reasons for this can be many (e.g., response differences of laboratory animal species, experiment design, etc.). Nevertheless, this type of discrepancy doesn’t override the mission of the USEPA—to protect the public health. Given current data for these compounds, it’s likely the agency will focus on studies reporting adverse effects. This is because those studies showing no adverse effects are not sufficient to counter the studies where adverse effects are seen. Given this situation, there appears to be a continuing interest by USEPA scientists since a recent abstract1 reported sodium chlorate causing adverse thyroidal effects at high doses (1 g/L or greater; approximate dose of 100 mg/kg-day) in drinking water for at least 90 days (although its authors concede: “This abstract does not reflect USEPA policy”).

Uncertainty factors
So, even if the database is small, contradictory or incomplete, an RfD or surrogate toxicity guideline value can be established. To develop an RfD, one divides the dose at the identified critical endpoint by uncertainty factors ranging from one to 10,000. Uncertainty factors are values of 10 that account each for: 1) a small database, 2) developing an RfD from a dose for a lowest observable adverse effect level (LOAEL) rather than a NOAEL, 3) inter-species extrapolation, and 4) intra-species sensitivity. For example, if we assume a 10 mg/kg-day LOAEL dose and apply uncertainty factors of 1,000 (10 × 10 × 10), the resulting hypothetical RfD would be 0.01 mg/kg-day. With the assumption of an average person weighing 70 kg and consuming 2 liters of water per day, this would translate into a SDWL of 0.35 ppm. The engineering measures needed to comply with this level may be costly and overly conservative from a protection of the public health point.

What could happen to the RfD if additional animal studies were conducted? One would design studies to establish a NOAEL for the critical endpoint. For example, if a NOAEL was established to be 7 mg/kg-day, then the hypothetical RfD would be 0.07 mg/kg-day (this includes a lower uncertainty factor of 100 because of the use of a NOAEL), which results in a SDWL of 2.5 ppm. The additional data increased the SDWL by a factor of slightly over 7 and is protective of the public health. If you compare the cost of conducting experiments to the cost of compliance, it might make economic sense to conduct the studies. Although MCLs and MRDLs aren’t derived in the same manner as an RfD, there are similarities (e.g., look for doses with no adverse effects) and there’s momentum toward standardizing toxicity guideline values—perchlorate being a present-day example.

Manufacturers and users of chlorine dioxide should keep a watchful eye on future research initiatives with regard to the investigation of the toxicity of chlorine dioxide and its DBPs. Also, toxicologists should be engaged who are familiar with USEPA’s method of establishing toxicity guideline values to assist in setting strategies. When setting strategies, it is critical a safe dose be identified that’s protective of the most sensitive members of the public. This can be accomplished by either establishing a toxicity guideline value that’s overly conservative or one that is more accurate. The toxicity guideline value can have a huge economic impact on manufacturers and users of chlorine dioxide if the conservative approach is employed. Monitoring research and engaging the expertise of a toxicologist early saves money because developing strategies based on the use of data allows for making proactive economic decisions based on a more accurate scientific approach. Otherwise, when an issue ignites, considerable resources are used to respond—mostly in a defensive and less cost-effective manner.


  1. Hooth, M.J., et al., “Sodium chlorate treatment results in a dose-dependent increase in rat thyroid follicular cell hyperplasia following subchronic exposure in drinking water,” The Toxicologist, 54:271, March, 2000.
  2. USEPA, Alternative Disinfectants and Oxidants Guidance Manual (EPA/815/R-99/014), Office of Water, Washington, D.C., 1999.
  3. USEPA, Glossary of IRIS Terms, revised October 1999:

About the author
Richard Pleus is principal and senior toxicologist for Intertox, a scientific public health consultant. Intertox provides expertise in toxicology, environmental science, epidemiology and scientific research that enables it to provide service in the areas of risk assessment and communication, exposure analysis, toxicological evaluation and information research. Dr. Pleus can be contacted at (206) 443-2115, (206) 443-2117 (fax) or email: rcpleus@intertox.comor website:

Kids and the Modes of Action
New initiatives by the USEPA are having a greater impact on the process of determining toxicity guideline values. Two of these initiatives focus on an environmental chemical’s “mode of action” and its impact on “children’s health.” The characterization of a chemical’s mode of action is defined as identifying the key physiological events leading to the development of adverse effects. For those chemicals without a large toxicological database, chemicals with a similar mode of action are thought to have similar toxicological effects. Perchlorate’s mode of action is inhibition of iodide uptake into the thyroid gland. Thus, at high doses, it reduces the production of thyroid hormones. The mode of action for chlorine dioxide and its DBPs was thought to be that chlorine dioxide would undergo hydrolysis to form chlorite (ClO2-) and chlorate (ClO3-) ions then—like perchlorate—inhibit the uptake of iodide into the thyroid gland. However, animal studies have provided evidence that this is unlikely and the specific mechanism has yet to be determined.

In addition, USEPA policy is currently focused on protecting children. The U.S. government is making children’s health concerns in relation to environmental chemicals a priority. Federal Order 13045 (EO 1997) requires “each federal agency shall make it a high priority to identify and assess environmental health and safety risks that may disproportionately affect children, and shall ensure that their policies, programs, and standards address disproportionate risks that result from our environmental health risks or safety risks.” This assessment addresses the possibility that children, infants and the developing fetus may be more susceptible due to a potential for increased formation and distribution of potentially toxic and carcinogenic metabolites in their systems.

For the most current information, the USEPA Office of Ground Water and Drinking Water’s Microbials and Disinfection By-Products Rule webpage is located at:


Comments are closed.