Under the Safe Drinking Water Act, the U.S. Environmental Protection Agency (EPA) is required to establish the concentrations of contaminants that are permitted in public drinking-water systems. A public water system is defined by EPA as a “system for the provision to the public of water for human consumption through pipes or other constructed conveyances, if such system has at least fifteen service connections or regularly serves at least twenty-five individuals” (63 Fed. Reg. 41940 ). Section 1412 of the act, as amended in 1986, requires EPA to publish maximum-contaminant-level goals (MCLGs) and promulgate national primary drinking-water regulations (maximum contaminant levels [MCLs]) for contaminants in drinking water that might cause any adverse effect on human health and that are known or expected to occur in public water systems. MCLGs are health goals set at concentrations at which no known or expected adverse health effects occur and the margins of safety are adequate. MCLGs are not regulatory requirements but are used by EPA as a basis for establishing MCLs. MCLs are enforceable standards to be set as close as possible to the MCLG with use of the best technology available. For some contaminants, EPA also establishes secondary maximum contaminant levels (SMCLs), which are nonenforceable guidelines for managing drinking water for aesthetic, cosmetic, or technical effects related to public acceptance of drinking water.
Fluoride is one of the natural contaminants found in public drinking water supplies regulated by EPA. In 1986, an MCLG of 4 milligrams per liter (mg/L) and an SMCL of 2 mg/L were established for fluoride, and an MCL of 4 mg/L was promulgated. It is important to make the distinction that EPA’s standards are guidelines for restricting the amount of naturally
occurring fluoride in drinking water; they are not recommendations about the practice of adding fluoride to public drinking-water systems (see below). In this report, the National Research Council’s (NRC’s) Committee on Fluoride in Drinking Water reviews the nature of the human health risks from fluoride, estimates exposures to the general public from drinking water and other sources, and provides an assessment of the adequacy of the MCLG for protecting public health from adverse health effects from fluoride and of the SMCL for protecting against cosmetic effects. Assessing the efficacy of fluoride in preventing dental caries is not covered in this report.
This chapter briefly reviews the sources of fluoride in drinking water, states the task the committee addressed, sets forth the committee’s activities and deliberative process in developing the report, and describes the organization of the report.
FLUORIDE IN DRINKING WATER
Fluoride may be found in drinking water as a natural contaminant or as an additive intended to provide public health protection from dental caries (artificial water fluoridation). EPA’s drinking water standards are restrictions on the amount of naturally occurring fluoride allowed in public water systems, and are not recommendations about the practice of water fluoridation. Recommendations for water fluoridation were established by the U.S. Public Health Service, and different considerations were factored into how those guidelines were established.
Fluoride occurs naturally in public water systems as a result of runoff from weathering of fluoride-containing rocks and soils and leaching from soil into groundwater. Atmospheric deposition of fluoride-containing emissions from coal-fired power plants and other industrial sources also contributes to amounts found in water, either by direct deposition or by deposition to soil and subsequent runoff into water. Of the approximately 10 million people with naturally fluoridated public water supplies in 1992, around 6.7 million had fluoride concentrations less than or equal to 1.2 mg/L (CDC 1993). Approximately 1.4 million had natural fluoride concentrations between 1.3 and 1.9 mg/L, 1.4 million had between 2.0 and 3.9 mg/L, and 200,000 had concentrations equal to or exceeding 4.0 mg/L. Exceptionally high concentrations of fluoride in drinking water are found in areas of Colorado (11.2 mg/L), Oklahoma (12.0 mg/L), New Mexico (13.0 mg/L), and Idaho (15.9 mg/L).
Areas of the United States with concentrations of fluoride in drinking water greater than 1.3 mg/L are all naturally contaminated. As discussed
below, a narrow concentration range of 0.7 to 1.2 mg/L is recommended when decisions are made to intentionally add fluoride into water systems. This lower range also occurs naturally in some areas of the United States. Information on the fluoride content of public water supplies is available from local water suppliers and local, county, or state health departments.
Since 1945, fluoride has been added to many public drinking-water supplies as a public-health practice to control dental caries. The “optimal” concentration of fluoride in drinking water for the United States for the prevention of dental caries has been set at 0.7 to 1.2 mg/L, depending on the mean temperature of the locality (0.7 mg/L for areas with warm climates, where water consumption is expected to be high, and 1.2 mg/L for cool climates, where water consumption is low) (PHS 1991). The optimal range was determined by selecting concentrations that would maximize caries prevention and limit enamel fluorosis, a dose-related mottling of teeth that can range from mild discoloration of the surface to severe staining and pitting. Decisions about fluoridating a public drinking-water supply are made by state or local authorities. CDC (2002a) estimates that approximately 162 million people (65.8% of the population served by public water systems) received optimally fluoridated water in 2000.
The practice of fluoridating water supplies has been the subject of controversy since it began (see reviews by Nesin 1956; Wollan 1968; McClure 1970; Marier 1977; Hileman 1988). Opponents have questioned the motivation for and the safety of the practice; some object to it because it is viewed as being imposed on them by the states and as an infringement on their freedom of choice (Hileman 1988; Cross and Carton 2003). Others claim that fluoride causes various adverse health effects and question whether the dental benefits outweigh the risks (Colquhoun 1997). Another issue of controversy is the safety of the chemicals used to fluoridate water. The most commonly used additives are silicofluorides, not the fluoride salts used in dental products (such as sodium fluoride and stannous fluoride). Silicofluorides are one of the by-products from the manufacture of phosphate fertilizers. The toxicity database on silicofluorides is sparse and questions have been raised about the assumption that they completely dissociate in water and, therefore, have toxicity similar to the fluoride salts tested in laboratory studies and used in consumer products (Coplan and Masters 2001).
It also has been maintained that, because of individual variations in exposure to fluoride, it is difficult to ensure that the right individual dose to protect against dental caries is provided through large-scale water fluoridation. In addition, a body of information has developed that indicates
the major anticaries benefit of fluoride is topical and not systemic (Zero et al. 1992; Rölla and Ekstrand 1996; Featherstone 1999; Limeback 1999a; Clarkson and McLoughlin 2000; CDC 2001; Fejerskov 2004). Thus, it has been argued that water fluoridation might not be the most effective way to protect the public from dental caries.
Public health agencies have long disputed these claims. Dental caries is a common childhood disease. It is caused by bacteria that colonize on tooth surfaces, where they ferment sugars and other carbohydrates, generating lactic acid and other acids that decay tooth enamel and form a cavity. If the cavity penetrates to the dentin (the tooth component under the enamel), the dental pulp can become infected, causing toothaches. If left untreated, pulp infection can lead to abscess, destruction of bone, and systemic infection (Cawson et al. 1982; USDHHS 2000). Various sources have concluded that water fluoridation has been an effective method for preventing dental decay (Newbrun 1989; Ripa 1993; Horowitz 1996; CDC 2001; Truman et al. 2002). Water fluoridation is supported by the Centers for Disease Control and Prevention (CDC) as one of the 10 great public health achievements in the United States, because of its role in reducing tooth decay in children and tooth loss in adults (CDC 1999). Each U.S. Surgeon General has endorsed water fluoridation over the decades it has been practiced, emphasizing that “[a] significant advantage of water fluoridation is that all residents of a community can enjoy its protective benefit…. A person’s income level or ability to receive dental care is not a barrier to receiving fluoridation’s health benefits” (Carmona 2004).
As noted earlier, this report does not evaluate nor make judgments about the benefits, safety, or efficacy of artificial water fluoridation. That practice is reviewed only in terms of being a source of exposure to fluoride.
HISTORY OF EPA’S REGULATION OF FLUORIDE
In 1975, EPA proposed an interim primary drinking-water regulation for fluoride of 1.4-2.4 mg/L. That range was twice the “optimal” range of 0.7-1.2 mg/L recommended by the U.S. Public Health Service for water fluoridation. EPA’s interim guideline was selected to prevent the occurrence of objectionable enamel fluorosis, mottling of teeth that can be classified as mild, moderate, or severe. In general, mild cases involve the development of white opaque areas in the enamel of the teeth, moderate cases involve visible brown staining, and severe cases include yellow to brown staining and pitting and cracking of the enamel (NRC 1993). EPA considered objectionable enamel fluorosis to involve moderate to severe cases with dark stains and pitting of the teeth.
The history of EPA’s regulation of fluoride is documented in 50 Fed. Reg. 20164 (1985). In 1981, the state of South Carolina petitioned EPA
to exclude fluoride from the primary drinking-water regulations and to set only an SMCL. South Carolina contended that enamel fluorosis should be considered a cosmetic effect and not an adverse health effect. The American Medical Association, the American Dental Association, the Association of State and Territorial Dental Directors, and the Association of State and Territorial Health Officials supported the petition. After reviewing the issue, the U.S. Public Health Service concluded there was no evidence that fluoride in public water supplies has any adverse effects on dental health, as measured by loss of teeth or tooth function. U.S. Surgeon General C. Everett Koop supported that position. The National Drinking Water Advisory Council (NDWAC) recommended that enamel fluorosis should be the basis for a secondary drinking-water regulation. Of the health effects considered to be adverse, NDWAC found osteosclerosis (increased bone density) to be the most relevant end point for establishing a primary regulation.
EPA asked the U.S. Surgeon General to review the available data on the nondental effects of fluoride and to determine the concentrations at which adverse health effects would occur and an appropriate margin of safety to protect public health. A scientific committee convened by the surgeon general concluded that exposure to fluoride at 5.0 to 8.0 mg/L was associated with radiologic evidence of osteosclerosis. Osteosclerosis was considered to be not an adverse health effect but an indication of osseous changes that would be prevented if the maximum content of fluoride in drinking water did not exceed 4 mg/L. The committee further concluded that there was no scientific documentation of adverse health effects at 8 mg/L and lower; thus, 4 mg/L would provide a margin of safety. In 1984, the surgeon general concluded that osteosclerosis is not an adverse health effect and that crippling skeletal fluorosis was the most relevant adverse health effect when considering exposure to fluoride from public drinking-water supplies. He continued to support limiting fluoride concentrations to 2 mg/L to avoid objectionable enamel fluorosis (50 Fed. Reg. 20164 ).
In 1984, NDWAC took up the issue of whether psychological and behavioral effects from objectionable enamel fluorosis should be considered adverse. The council concluded that the cosmetic effects of enamel fluorosis could lead to psychological and behavioral problems that affect the over-all well-being of the individual. EPA and the National Institute of Mental Health convened an ad hoc panel of behavioral scientists to further evaluate the potential psychological effects of objectionable enamel fluorosis. The panel concluded that “individuals who have suffered impaired dental appearance as a result of moderate or severe fluorosis are probably at increased risk for psychological and behavioral problems or difficulties” (R. E. Kleck, unpublished report, Nov. 17, 1984, as cited in 50 Fed. Reg. 20164 ). NDWAC recommended that the primary drinking-water guideline for fluoride be set at 2 mg/L (50 Fed. Reg. 20164 ).
On the basis of its review of the available data and consideration of the recommendations of various advisory bodies, EPA set an MCLG of 4 mg/L on the basis of crippling skeletal fluorosis (50 Fed. Reg. 47,142 ). That value was calculated from an estimated lowest-observed-adverse-effect level of 20 mg/day for crippling skeletal fluorosis, the assumption that adult water intake is 2 L per day, and the application of a safety factor of 2.5. This factor was selected by EPA to establish an MCLG that was in agreement with a recommendation from the U.S. Surgeon General. In 1986, the MCL for fluoride was promulgated to be the same as the MCLG of 4 mg/L (51 Fed. Reg. 11,396 ).
EPA also established an SMCL for fluoride of 2 mg/L to prevent objectionable enamel fluorosis in a significant portion of the population (51 Fed. Reg. 11,396 ). To set that guideline, EPA reviewed data on the incidence of moderate and severe enamel fluorosis and found that, at a fluoride concentration of 2 mg/L, the incidence of moderate fluorosis ranged from 0% to 15%. Severe cases appeared to be observed only at concentrations above 2.5 mg/L. Thus, 2 mg/L was considered adequate for preventing enamel fluorosis that would be cosmetically objectionable. EPA established the SMCL as an upper boundary guideline for areas that have high concentrations of naturally occurring fluoride. EPA does not regulate or promote the addition of fluoride to drinking water. If fluoride in a community water system exceeds the SMCL but not the MCL, a notice about potential risk of enamel fluorosis must be sent to all customers served by the system (40 CFR 141.208).
In the early 1990s, the NRC was asked to independently review the health effects of ingested fluoride and EPA’s MCL. The NRC (1993) found EPA’s MCL of 4 mg/L to be an appropriate interim standard. Its report identified inconsistencies in the fluoride toxicity database and gaps in knowledge. Accordingly, the NRC recommended research in the areas of fluoride intake, enamel fluorosis, bone strength and fractures, and carcinogenicity. A list of the specific recommendations from that report is provided in Box 1-1.
The Safe Drinking Water Act requires that EPA periodically review existing standards for water contaminants. Because of that requirement and new research on fluoride, EPA’s Office of Water requested that the NRC reevaluate the adequacy of the MCLG and SMCL for fluoride to protect public health. The NRC assigned this task to the standing Committee on Toxicology, and convened the Committee on Fluoride in Drinking Water. The committee was asked to review toxicologic, epidemiologic, and clinical data, particularly data published since 1993, and exposure data on orally ingested fluoride from drinking water and other sources (e.g., food, tooth-
Recommendations from NRC (1993) Report
Intake, Metabolism, and Disposition of Fluoride
paste, dental rinses). On the basis of those reviews, the committee was asked to evaluate independently the scientific basis of EPA’s MCLG of 4 mg/L and SMCL of 2 mg/L in drinking water and the adequacy of those guidelines to protect children and others from adverse health effects. The committee was asked to consider the relative contribution of various fluoride sources (e.g., food, dental-hygiene products) to total exposure. The committee also was asked to identify data gaps and make recommendations for future research relevant to setting the MCLG and SMCL for fluoride. Addressing questions of economics, risk-benefit assessment, and water-treatment technology was not part of the committee’s charge.
The committee is aware that some readers expect this report to make a determination about whether public drinking-water supplies should be fluoridated. That expectation goes beyond the committee’s charge. As noted above, the MCLG and SMCL are guidelines for areas where fluoride con-
centrations are naturally high. They are designed with the intent to protect the public from adverse health effects related to fluoride exposure and not as guidelines to provide health benefits.
To accomplish its task, the committee held six meetings between August 2003 and June 2005. The first two meetings involved data-gathering sessions that were open to the public. The committee heard presentations from EPA, CDC, individuals involved in fluoride research, fluoridation supporters, and antifluoridation proponents. The committee also reviewed a large body of written material on fluoride, primarily focusing on research that was completed after publication of the 1993 NRC report. The available data included numerous research articles, literature reviews, position papers, and unpublished data submitted by various sources, including the public. Each paper and submission was evaluated case by case on its own merits.
Unless otherwise noted, the term fluoride is used in this report to refer to the inorganic, ionic form. Most of the nonepidemiologic studies reviewed involved exposure to a specified fluoride compound, usually sodium fluoride. Various units of measure are used to express exposure to fluoride in terms of exposure concentrations and internal dose (see Table 1-1 and Chapter 3). To the extent possible, the committee has tried to use units that allow for easy comparisons.
In this report, the committee updates information on the issues considered in the 1993 review—namely, data on pharmacokinetics; dental effects; skeletal effects; reproductive and developmental effects; neurological and behavioral effects; endocrine effects; gastrointestinal, renal, hepatic, and immune effects; genotoxicity; and carcinogenicity. More inclusive reviews are provided on effects to the endocrine and central nervous systems, because the previous NRC review did not give those effects as much attention. The committee used a general weight-of-evidence approach to evaluate the literature, which involved assessing whether multiple lines of evidence
TABLE 1-1 Units Commonly Used for Measuring Fluoride
ABBREVIATIONS: mg/kg, milligrams per kilogram; mg/L, milligrams per liter; µmol/L, micromoles per liter; ppm, parts per million.
indicate a human health risk. This included an evaluation of in vitro assays, animal research, and human studies (conducted in the United States and other countries). Positive and negative results were considered, as well as mechanistic and nonmechanistic information. The collective evidence was considered in perspective with exposures likely to occur from fluoride in drinking water at the MCLG or SMCL.
In evaluating the effects of fluoride, consideration is given to the exposure associated with the effects in terms of dose and time. Dose is a simple variable (such as mg/kg/day), and time is a complex variable because it involves not only the frequency and duration of exposure but also the persistence of the agent in the system (kinetics) and the effect produced by the agent (dynamics). Whether the key rate-limiting events responsible for the adverse effect are occurring in the kinetic or in the dynamic pathway is important in understanding the toxicity of a chemical and in directing future research (see Rozman and Doull 2000). The committee also attempts to characterize fluoride exposures from various sources to different subgroups within the general population and to identify subpopulations that might be particularly susceptible to the effects of fluoride.
STRUCTURE OF THE REPORT
The remainder of this report is organized into 10 chapters. Chapter 2 characterizes the general public’s exposure to fluoride from drinking water and other sources. Chapter 3 provides a description of the chemistry of fluoride and pharmacokinetic information that was considered in evaluating the toxicity data on fluoride. In Chapters 4-9, the committee evaluates the scientific literature on adverse effects of fluoride on teeth, the musculoskeletal system, reproduction and development, the nervous system, the endocrine system, the gastrointestinal system, the kidneys, the liver, and the immune system. Chapter 10 evaluates the genotoxic and carcinogenic potential of fluoride. Finally, Chapter 11 provides an assessment of the most significant health risks from fluoride in drinking water and its implications for the adequacy of EPA’s MCLG and SMCL for protecting the public.