by Paul Connett
It is important to put the role of the CDC in this matter in perspective. As we explained in chapter 4, there is only one division at the CDC involved with fluoridation, the Oral Health Division (OHD). This division is largely staffed with personnel with dental qualifications but very few with qualifications in general or specialized medicine. Nor do any appear to have qualifications in toxicology or risk assessment. There appears to be no individual scientist (or group of scientists), let alone a division, at the CDC with the responsibility to oversee the safety of water fluoridation. This is all left to the OHD, which has a clear conflict of interest in this matter, because it actively promotes water fluoridation.
While it is clear that political factors (i. e. , doing whatever is necessary to support a long-standing policy) appear to be influencing both the ADA’s and the CDC’s positions on the relevance of the NRC report to water fluoridation, it is important to stress the scientific flaws in their respective claims that somehow fluoride is bad at 4 ppm but okay at 1 ppm, because the same argument is being made by several other pro-fluoridation government agencies, spokespersons, and consultants. 25–27 We will examine the case of Bazian Ltd.
Bazian’s Critique
As part of its consultation process for the proposal to fluoridate Southampton and surrounding communities, the UK South Central Strategic Health Authority (SHA) held three public meetings in October and November 2008. At those meetings Paul Connett and others drew attention to the NRC report. In response, the SHA hired the consulting company Bazian Ltd. to review the report.
Bazian’s only substantive point was that the NRC panel was considering higher concentrations of fluoride than those used for fluoridation: “The maximum levels of water-borne fluoride for cosmetic or health safety recommended by the EPA, and examined in the book [the NRC report], are two to four times higher than the level of water fluoridation proposed for Southampton and higher than the 1. 5ppm maximum level laid down by the European Union. ”28
Using this as a reason, SHA dismissed the NRC report without referring to a single page, let alone a single one of the studies referenced. The following statement appeared in the final SHA report, authored by Professor John Newton, endorsing fluoridation for Southampton: “The SHA commissioned the. . . specialist expert organization (Bazian) to undertake a critical appraisal of the National Research Council report. Their conclusion was that the report is not relevant to the proposed scheme because it considers much higher levels of fluoride (4ppm) than those envisaged in Southampton (1ppm). ”29
Bazian’s actual wording in its summary was as follows: “The question it [the report] examines does not have direct relevance to the issue of fluoridation in Southampton” [emphasis added]. 30 In other words, Bazian did not deny the obvious fact that the NRC data were relevant but found a way to imply that the relevance was “indirect, ” whatever difference that may make. Although Bazian’s words were quoted in the SHA paper, the word “direct” was omitted in Professor Newton’s text. For whatever reason, Newton felt free to ignore the NRC report when recommending fluoridation in Southampton—an action that many may deem incautious or even reckless.
It is difficult to know just what Bazian meant to imply with the word “direct” and what the firm’s representatives would have said if asked to explain what “indirect” evidence the NRC review had provided. Had they been challenged, they might have been forced to acknowledge the crucial difference between concentration and dose that we discuss in chapter 1 (someone drinking 2 liters of water with a concentration of 1 ppm would get the same dose of fluoride as someone drinking 1 liter of water at 2 ppm). They might have admitted that what was needed here was an analysis that would establish that there was an adequate margin of safety sufficient to protect everyone in the Southampton area from the adverse health effects documented in the NRC review (see chapter 20). Such an exercise would have required consideration of the doses at which these effects were observed, and the expected range of dose experienced by a population whose consumption of water and of fluoride obtained from other sources was uncontrolled. It would have also required choosing an adequate safety margin to account for the range of sensitivity to any toxic substance expected across a large population (see chapters 1, 13 and 20). Establishing safety under these vital requirements would have been difficult to do, especially since in chapter 2 of the NRC review, the panel established that certain subsets of the American population are exceeding the EPA’s IRIS reference dose of 0. 06 mg/kg/day31 while drinking fluoridated water at 1 ppm, including infants who are fed formula made with tap water.
Here are just two of several other pieces of indirect evidence that needed consideration. The NRC reviewed studies that indicated an association between a tripling of hip fractures in elderly people drinking water at 4. 3 ppm, and a possible doubling at 1. 5 ppm, in a relatively small study population32 (see chapter 17). It is difficult to see how Bazian or the SHA could justify exposing a much larger population to fluoride at 1 ppm over a lifetime, especially since that population would include some people with impaired kidney function. Similar considerations apply to the association between altered thyroid function and people drinking water with fluoride at 2. 3 ppm. 33 Such considerations are discussed further in chapter 20.
Kathleen Thiessen, PhD, an NRC panel member, responded to the claim that the NRC report is not relevant to public water fluoridation in the course of two statements, 34, 35 of which excerpts follow:
Scope of the NRC report. The National Research Council’s committee on fluoride in drinking water was asked to review the adequacy of EPA’s Maximum Contaminant Level Goal (MCLG) of 4 mg/L fluoride in drinking water and the corresponding Secondary Maximum Contaminant Level (SMCL) of 2 mg/L. The committee concluded that those regulatory limits are not protective of public health. The committee was not asked to review the safety of so-called “optimal” concentrations of fluoride in drinking water (0. 7–1. 2 mg/L, as used in deliberate fluoridation of public drinking water supplies), although much of the report is relevant to such a review (discussed further below). In addition, the committee was not asked to review the efficacy or reported benefits of fluoridation, on the basis of which community water fluoridation was instituted. The committee also did not review in any detail either the history or the politics of water fluoridation.
Relevance of the NRC report to water fluoridation. Although the NRC report did not examine the safety, efficacy, or benefits of water fluoridation, or specifically evaluate the toxic effects of “optimal” levels of water fluoride on humans, the committee did examine a number of issues that are relevant to such evaluations. In particular, the committee did an extremely thorough review of fluoride intake in the U. S. , by age group, considering all sources of fluoride intake (water, dentifrices, food, air, soil, pesticides, pharmaceuticals), including fluoridated drinking water. In addition, the committee looked specifically at population subgroups of special concern, for example, due to very high water consumption or to impaired fluoride excretion. A number of the toxicity studies that the committee reviewed involved fluoridated water or exposures equivalent to those expected with fluoridated water. . .
Hazards of fluoride exposure. The NRC report concluded that the existing MCLG of 4 mg/L is not protective of human health. This conclusion was based largely on health effects that have long been considered specific to fluoride and significant enough to warrant protection, namely dental fluorosis and skeletal fluorosis. The NRC’s review differed from previous reviews of fluoride by saying that severe dental fluorosis is an adverse health effect (not merely a cosmetic effect), that stage II as well as stage III skeletal fluorosis is an adverse health effect, and that a fluoride concentration of 4 mg/L is likely not protective with respect to an increased risk of bone fracture. The NRC report indicated that at 2 or 4 mg/L, bone fluoride concentrations can reach the ranges historically associated with stage II and III skeletal fluorosis. The committee was not able to rule out a carcinogenic effect of fluoride or of “water fluoridat
ion” (i. e. , due to some substance added along with an impure fluoridating agent). Nor was the committee able to rule out the possibility that fluoridation is associated with an increased risk of Down syndrome in children of young mothers. The committee also reported that fluoride exposure is plausibly associated with a number of other health effects, including neurotoxicity, gastrointestinal problems, and endocrine problems, and that even though these effects are not necessarily specific to fluoride exposure, the associations cannot be ruled out and need further study.
With respect to dental fluorosis, skeletal fluorosis, and risk of bone fracture, the NRC committee considered primarily studies in which populations were exposed to concentrations of fluoride in drinking water of around 4 mg/L; from those studies the committee concluded that 4 mg/L is not protective of those effects. The committee did not, for any endpoint, determine a “no-effect level, ” an individual intake level (mg per day of fluoride intake per kg body weight) below which no adverse health effects occur. However, the ranges of intake levels, or estimated average intake levels, associated with a number of adverse effects, are in the range of intakes expected with fluoridated drinking water in the U. S. Fluoride exposures in the U. S. are driven largely by consumption of drinking water and beverages made with tap water. Water intake for a given age group varies substantially— around a factor of 100 between the highest and lowest consumption rates (discussed in the NRC report). The result of this is that for water fluoride at 1 mg/L vs. water fluoride at 4 mg/L, there will be a huge overlap between the respective populations, with apparent differences only at the very highest water intakes. In other words, any effect seen at 4 mg/L is probably going to occur in some people at 1 mg/L (e. g. , in the people with highest water consumption or in people with impaired fluoride excretion), but this might easily be missed in the sample sizes typically used in studies. 36
Carcinogenicity. Chapter 10 of the NRC report also reviewed human and animal studies of carcinogenicity, in addition to genotoxicity studies, although the NRC’s review did not include a number of the older studies. . . The committee unanimously concluded that “Fluoride appears to have the potential to initiate or promote cancers, ” even though the overall evidence is “mixed. ” Referring to the animal studies, the committee also said that “the nature of uncertainties in the existing data could also be viewed as supporting a greater precaution regarding the potential risk to humans. ” The committee also discussed the limitations of epidemiologic studies, especially ecologic studies (those in which group, rather than individual, measures of exposure and outcome are used), in detecting small increases in risk—in other words, the studies are not sensitive enough to identify small increases in cancer risk; therefore a “negative” study does not necessarily mean that there is no risk. . .
While the NRC committee did not assign fluoride to a specific category of carcinogenicity (i. e. , known, probable, or possible), the committee did not consider either “insufficient information” or “clearly not carcinogenic” to be applicable. The committee report includes a discussion of how EPA establishes drinking water standards for known, probable, or possible carcinogens; such a discussion would not have been relevant had the committee not considered fluoride to be carcinogenic. The question becomes one of how strongly carcinogenic fluoride is, and under what circumstances. As mentioned, fluoride may be a cancer promoter rather than an initiator, although the two mechanisms are not mutually exclusive. 37
Basis for establishing fluoride concentrations in local drinking water supplies. Historically, the local temperature (the “annual average of maximum daily air temperatures” over a minimum of 5 years) has been used as the basis for recommending a given level of fluoride in the drinking water (e. g. , CDC 1995). In practice (reviewed by the NRC), there seems to be little difference in water consumption for many people with temperature, season, or location. Obviously, for people with high levels of activity, water consumption can be very high. At present, basketball players or gymnasts, for example, will probably have similar rates of water consumption no matter which state they live in; however, under current guidelines, some of them will have water with 0. 7 mg/L fluoride, while others will have water with 1. 2 mg/L. Also, most states do not appear to account for temperature variations within a state, such that the water fluoridation levels are the same for the colder and hotter parts of the same state.
Concerns about silicofluorides. A number of issues have been raised concerning the use of silicofluorides as the fluoridating agent in most public water supplies (discussed briefly in the NRC report). These include increased lead in children’s blood, increased leaching of lead into water from plumbing fixtures, and the addition of other substances to the drinking water along with the silicofluorides. For instance, the MCLGs for arsenic and lead are 0, based on health risks; however, the actual level permitted (the Maximum Contaminant Level, or MCL) is above 0 (to account for difficulty in removing it or in measuring it). However, in the addition of the impure silicofluorides to drinking water, some arsenic and lead are generally added as well, although the resulting concentration must stay below the MCL. Given that the MCLGs are 0, the obvious question is whether knowingly adding any amount, however tiny, is appropriate. 38
Summary
The 2006 National Research Council report was the first U. S. report to look at low-level fluoride toxicity in a balanced way. The reporting panel’s task was to determine whether the maximum contaminant level goal for drinking water, currently 4 ppm, was appropriate for protecting health. The report concluded that the MCLG was too high and should be reduced. The report is clearly relevant to fluoridation since, if 4 ppm is too high (by an unspecified amount) to be acceptable as a contaminant, it cannot be sensible to deliberately add 1 ppm. That implies a safety margin of less than four times, possibly much less—absurdly small by normal toxicological standards. Despite this, the major promoters of fluoridation hastened to state that the report was irrelevant to fluoridation and could be completely disregarded, on the spurious grounds that it dealt only with exposure to fluoride at more than 2 ppm. Acceptance of such a tiny margin of safety indicates a cavalier disregard for public health. The report identified three main concerns—stage II skeletal fluorosis, bone fractures, and severe dental fluorosis—but also drew attention to other potential health hazards especially to the endocrine system and the developing brain.
In the following chapters we review the impacts of fluoride on various tissues of concern and buttress our comments with citations from the NRC review, the most comprehensive text on fluoride’s toxicology available to date. However, several important health studies have been published or translated into English since the NRC report was published in 2006, and they add further weight to the NRC findings. These studies include many more research papers on fluoride’s impact on the brain (chapter 15) as well as Bassin et al. ’s important paper on osteosarcoma39 (chapter 18).
•
15 •
Fluoride and the Brain
Nowhere do the dangers of exposing a whole population to uncontrolled doses of a toxic substance become more apparent than in the possibility that they may damage brain function. It has been documented time and again that very small amounts of some chemical substances can cause subtle yet significant changes in a child’s mental development and behavior. Once a small chemical change occurs over a wide and varied population, there may be unintended and unfortunate consequences, not only for the individual but for the entire population. Such changes might not be recognized until the damage becomes irreversible or irreparable. This appears to have happened in the case of the impact of airborne lead on children’s IQ in the United States, when lead was present in gasoline. It doesn’t take a very large downwind shift in IQ in the whole population to have a dramatic effect in decreasing the number of intellectually gifted children and increasing the number of mentally handicapped.
The story of fluoride’s possible interaction with the brain began independently on at least three contine
nts:
• The use of uranium hexafluoride to separate uranium isotopes exposed atomic bomb workers to large amounts of fluoride. In the 1940s, the late Harold Hodge, working as chief toxicologist for the Manhattan Project, asked the head of the medical section, Colonel Stafford Warren, for funding to do an animal experiment to investigate the possible impact of fluoride on the brain. 1 However, the results of such an experiment would not emerge until almost fifty years later.
• In the late 1980s, Dr. Phyllis Mullenix, a toxicologist, developed a novel technique for exploring the possible changes to animal behavior that a chemical might cause. When she was hired by Forsyth Dental Center in Boston and asked to use that technique to analyze the neurotoxicity of the chemicals used in dentistry, she had little idea of how much trouble her discoveries would provoke in her life and career.
• When, in 1994, the New Zealand doctor Bruce J. Spittle, who specialized in psychiatric medicine, published a review speculating on the potential for fluoride to damage the brain, he probably had little idea how his work would become part of the findings that would ultimately challenge the practice in which his dental colleagues so firmly believed. 2 Spittle taught at Otago University, which housed the only dental school in the country.
• In the late 1980s, Chinese researchers found that children had lower IQs in villages in areas with high natural levels of fluoride in the water. Those researchers also probably had little notion of how their work— when it was eventually translated into English—might threaten the whole edifice of artificial water fluoridation.
Gradually, over the last twenty years or so, these strands have come together.
