Fluoride in water: A UK perspective

Journal of Fluorine Chemistry 126 (2005) 1448–1456 www.elsevier.com/locate/fluor

Review

Fluoride in water: A UK perspective Paul T.C. Harrison * MRC Institute for Environment and Health, University of Leicester, 94 Regent Road, Leicester LE1 7DD, UK Received 21 January 2005; received in revised form 30 August 2005; accepted 6 September 2005 Available online 2 November 2005

Abstract Fluoride occurs naturally in soil, water, plants and animals in trace quantities. When fluoride is ingested, some is taken up by body tissues, with long-term deposition in teeth and bones. Following the demonstration of a significant reduction in dental caries in childhood within populations exposed to higher levels of fluoride in drinking water, between 1964 and 1975 several Local Authority water fluoridation schemes were introduced in England and Wales, whereby the fluoride content was artificially increased to a level of 1 ppm (1 mg L 1). Although evidence continues to support the premise that fluoride in water helps protect children’s teeth against caries, there are a number of potential adverse impacts, notably dental fluorosis (mottling of teeth). The situation is complicated by the fact that many individuals receive additional exposure to fluoride through the use of fluoride toothpaste, for example. Nonetheless, fluoridation of water continues to be generally regarded as a safe, simple and cost-effective public health measure to reach children most at risk and reduce the incidence of dental caries. Available evidence on risk of hip and other bone fractures suggests no effect of fluoride in water, although a small percentage change (in either direction) cannot be ruled out. There appears to be no link between water fluoridation and either cancer in general or any specific cancer type, but an updated analysis of UK data on fluoridation and cancer rates has nonetheless been recommended. Evidence for additional health outcomes suggested by some to be associated with fluoride ingestion, and on other concerns related to the chemicals that are added during the fluoridation process and indirect effects such as increased leaching of lead from pipes and aluminium from cooking utensils, is weak but the area deserves to be kept under review. # 2005 Elsevier B.V. All rights reserved. Keywords: Dental health; Drinking water; Exposure; Fluoride; Fluoridation; Health effects

Contents 1. 2. 3.

Introduction . . . . . . . . . . . . . . . . . . . Fluoride exposure and absorption. . . . . Health impacts . . . . . . . . . . . . . . . . . 3.1. Dental caries . . . . . . . . . . . . . . 3.1.1. Effects of social class . . 3.2. Dental fluorosis . . . . . . . . . . . . 3.3. Bone health . . . . . . . . . . . . . . . 3.4. Cancer . . . . . . . . . . . . . . . . . . 3.5. Other health effects . . . . . . . . . 3.5.1. Immunological effects. . 3.5.2. Effects on reproduction . 3.5.3. Birth defects . . . . . . . . 3.5.4. Renal effects . . . . . . . . 3.5.5. Gastrointestinal tract . . . 3.5.6. Intelligence . . . . . . . . . 3.5.7. Thyroid (goitre) . . . . . . 3.5.8. Miscellaneous effects . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

1449 1450 1451 1451 1451 1452 1452 1452 1453 1453 1453 1453 1453 1454 1454 1454 1454

* Present address: Institute of Environment and Health, Cranfield University, Silsoe, Bedfordshire MK 45 4DT, UK. Tel.: +44 1525 863327; fax: +44 1525 863344. E-mail address: [email protected]. 0022-1139/$ – see front matter # 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.jfluchem.2005.09.009

P.T.C. Harrison / Journal of Fluorine Chemistry 126 (2005) 1448–1456

4.

3.6. Indirect effects of adding fluoride to water . . . . . . . 3.7. Substances added during the fluoridation process . . . 3.8. Dietary exposure to metals . . . . . . . . . . . . . . . . . . 3.9. Effects on bioavailability or toxicity of toxic metals . Conclusions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . .

. . . . . . .

1. Introduction Fluoride occurs naturally in soil, water, plants and animals in trace quantities. In groundwater, natural fluoride concentrations range from trace quantities to over 25 mg L 1. When fluoride is ingested by humans and other animals some is taken up by body tissues, with long-term deposition in teeth and bones. During the 1930s, it was discovered that children living in areas where drinking water contained naturally elevated levels of fluoride experienced less tooth decay.1 Subsequent epidemiological studies, notably those in the USA and Canada, established the benefits to dental health from drinking water artificially fluoridated at a level of 1 ppm (1 mg L 1)2 [3]. Fluoridation schemes began to be introduced into the United States in the 1940s, so that now about two-thirds of the United States population served by public water supplies, some 170 million people, consume fluoridated water [4]. In 1952, on the recommendation of the Medical Research Council (MRC), the British Government initiated a study into water fluoridation with a view to advising whether fluoride should be added to drinking water supplies in the UK [5,6]. As a result, several Local Authority water fluoridation schemes were introduced in England and Wales between 1964 and 1975, and some 6 million people now receive water in which the fluoride content has been artificially increased to a level of 1 ppm [3]. Major schemes are in operation in Birmingham and throughout the West Midlands, and also in Tyneside. In addition, about 500,000 people in this country receive water, which naturally contains fluoride at or about the level of 1 ppm. A further 1 million people receive water which naturally contains fluoride at a lower level, but which is still considered to confer some dental benefits [7]. In regions where levels of naturally occurring fluoride are high enough to cause serious adverse health effects (see below), fluoride content of drinking water is lowered. Various processes have been developed to do this [8]. 1 An account of the interesting history of how observations on the incidence of mottled teeth led to the discovery of a beneficial impact of fluoride on tooth decay can be found in Morbidity and Mortality Weekly Report, vol. 48, pp. 933– 993 (1999) (available at: http://www.cdc.gov/mmwr/preview/mmwrhtml/ mm4841lal.htm). 2 According to Hamilton [1] the concentration of fluoride in water required for prevention of dental caries is generally agreed to be in the range 0.7– 1.2 mg L 1, although the 1994 WHO report stated that ‘‘the world optimum concentration would normally be in the range 0.5–1.0 mg L 1’’. In the UK, concentrations of 0.3–0.7 mg L 1 are considered to afford below optimal protection against tooth decay, and at less than 0.3 mg L 1 it is said to be doubtful whether any benefit is gained [2].

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

1449

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

1454 1454 1455 1455 1455 1455 1455

In 1993, concerns were raised in the UK about a possible relationship between fluoridation of public drinking water and the incidence of osteoporosis. In response, an MRC Working Group was formed to review the evidence for this effect, and to advise whether further studies were required. The Working Group concluded that the benefits of water fluoridation appear to far outweigh the risks. However, they suggested that additional research be conducted to improve knowledge and to underpin public health policy [9]. In September 2000, the National Health Centre (NHS) Centre for Reviews and Dissemination at the University of York published a systematic review of epidemiological studies on water fluoridation and health [10]. The review had been commissioned by the Chief Medical Officer of the Department of Health (DH) in recognition of the fact that much of the research evidence linking water fluoridation to improved dental health had been undertaken several decades earlier (see ‘Saving Lives: Our Healthier Nation’, paragraph 9.20) [11]. The York Review confirmed some beneficial effects of water fluoridation on dental caries, but also suggested that this should be considered alongside the increased prevalence of dental fluorosis. Another key conclusion of the review was that little high quality research had been undertaken in the area of fluoride and health more broadly. The available research evidence was considered insufficient to allow a confident estimate of the risks that might be associated with non-dental health outcomes or of the potential benefit of water fluoridation on health inequalities associated with dental caries. A Working Group of the Medical Research Council was then established to take forward the conclusions and recommendations of the York Review and consider what further research might be required to improve the evidence base in the area of fluoride and health. The report by this Working Group on Water Fluoridation and Health [12], widely referenced in this article, identified areas of uncertainty regarding the balance between the benefits and disadvantages of water fluoridation and made recommendations for research to remove or reduce these uncertainties. The practice of fluoridating water has been endorsed by (inter alia) the World Health Organization, the British Medical Association, the Faculty of Public Health Medicine, the British Dental Association and, in the USA, by the Surgeon-General, the American Medical Association and the American Dental Association [12]. Proponents consider it a safe, simple and costeffective public health measure to reduce the incidence of dental caries [1]. Certainly it has been argued that, if the community has a piped water supply, water fluoridation is the most cost-effective method of reaching the whole population,

1450

P.T.C. Harrison / Journal of Fluorine Chemistry 126 (2005) 1448–1456

including children at highest risk [13]. However, there are potential health disadvantages of fluoride that have to be considered, including dental and skeletal fluorosis and other postulated detrimental effects, including immunotoxicity, carcinogenicity, genotoxicity, reprotoxicity, teratogenicity, renal toxicity and gastrointestinal tract toxicity. Also there are various additional potential sources of exposure to fluoride, other than drinking water. This article focuses on some of these key issues. 2. Fluoride exposure and absorption The major sources of ingested fluoride are water and other dietary sources (including tea, food and drinks prepared with fluoridated water), fluoride tablets and drops and other supplements (more relevant in the USA and elsewhere than in the UK), and unintentional ingestion of fluoridated toothpaste (especially by children) or fluoride mouth-rinses. In addition, some individuals are exposed to fluoride occupationally, or receive therapeutic doses of fluoride (for example, in the treatment of otosclerosis and osteoporosis) [12,14]. Some estimates, for adults and children, of fluoride ingestion from drinking water, tooth brushing and diet are shown in Table 1. Up until the 1960s, the ingestion of fluorides from water (whether natural or supplemented) probably represented the bulk of fluoride exposure for both adults and children in most populations [2].3 Since then, however, the availability of fluoride from other sources has changed significantly, and fluoride in drinking water is now recognised as just one component of an individual’s total fluoride intake. For example, in the 1970s fluoride started to be added to toothpastes and by 1978 96% of toothpaste on the market contained fluoride, usually at a concentration of 1000–1500 ppm (though lower fluoride toothpastes containing about 500 ppm fluoride are now available for use by children) [12]. Nonetheless, in UK fluoridated areas, drinking water probably remains the most important source of fluoride intake. A number of studies, using a variety of techniques, have attempted to estimate the mean daily intake of fluoride. Most have found intakes of 0.01–0.13 mg kg 1 body weight (bw), with most mean intake values between 0.03 and 0.04 mg kg 1 bw in non-fluoridated areas and 0.04– 0.06 mg kg 1 bw in fluoridated areas. However, individual intakes in children can exceed greatly the mean value, owing to ingestion of dentifrice, for example [15]. When a fluoride compound is ingested, some of the fluoride ion is retained in the mouth and incorporated into the teeth by ion exchange; but most of the fluoride is absorbed rapidly as non-ionised hydrogen fluoride (?) from the stomach and small intestine. A number of variables influence the amount of fluoride absorbed, including age, type of fluoride compound, pH conditions, and the concentration of magnesium, calcium and other ions [1,16]. When a readily soluble compound, such 3 With two exceptions: workers in industries handling fluoride concentrates, and osteoporosis patients given high fluoride doses therapeutically.

Table 1 Daily fluoride consumption from drinking water, tooth brushing and diet Source

Concentration/ content 1 b

Drinking water (mg L ) Tooth-brushing and mouth washing (mg) Diet (mg) Total intake

Intake (mg kg 1) a Child

Adult

1.2 0.145–0.66

0.084 0.033

0.034 0.002

0.2–0.4

0.01

0.006

0.127

0.042

Adapted from Hamilton [1]. a Assuming child and adult weights of 20 and 70 kg, respectively, together with 100% absorption. b Assumes child and adult water consumption of 1.4 and 2l day (likely to be an overestimate for the present day).

as sodium fluoride is ingested with water, nearly all the fluoride may be absorbed, but if the fluoride is taken with milk or food then the degree of absorption is reduced because of the formation of insoluble complexes or precipitates [16]. Blood plasma concentrations of fluoride usually peak 30– 60 min after ingestion. Fluoride is distributed to both hard and soft tissues, particularly bone and the kidneys. In adults and children, as little as 10% and as much as 50% or more, respectively, of ingested fluoride may be retained; the remainder is excreted, predominantly in urine but also in faeces and sweat [1]. Of the retained fluoride, approximately 97–99% becomes associated with bone and other calcified tissues [16]. However, the fluoride of calcified tissues is not irreversibly bound, and if fluoride intake is reduced over a long enough time-course, the concentrations in bone, etc. will decline through the mobilisation of fluoride by ion exchange. In addition, fluoride will be released by the process of bone resorption. Within the soft tissues of the body, fluoride rapidly establishes a steady state distribution between extracellular and intracellular fluids [16]. The question of the bioavailability of ingested fluoride is important, especially with respect to the possible influence of water hardness on uptake and differences between naturally fluoridated and artificially fluoridated water. Inorganic ions in the water certainly can interfere with fluoride absorption, but at the 1 ppm fluoride level this interference is considered to be biologically insignificant in normally composed drinking water. Only at high concentrations of calcium, magnesium and aluminium ions is fluoride absorption effectively reduced, owing to formation of less soluble fluoride complexes [17]. Considering possible differences in bioavailability between naturally fluoridated and artificially fluoridated water, Cremer and Buttner [17] concluded that ‘‘fluorides that either occur naturally in water or are added to communal supplies. . . to increase the fluoride level to 1 ppm F, yield fluoride ions which are almost completely absorbed from the gastrointestinal tract’’. In a more recent review, Jackson et al. [18] confirmed that, in terms of chemistry, there is no difference in bioavailability between added and ‘natural’ fluoride and that the impact of water hardness on the uptake of fluoride would be very small. Research commissioned by the UK Department of Health, as a direct result of the recommendation by the MRC

P.T.C. Harrison / Journal of Fluorine Chemistry 126 (2005) 1448–1456

Working Group [12], has tested this biologically in human volunteers.4 The specific aims of the study were ‘‘to compare the bioavailability of fluoride in naturally fluoridated water with artificially fluoridated water and to investigate the effect of water hardness on bioavailability of fluoride in drinking water’’. Blood fluoride concentrations were measured against time following the ingestion of four waters (naturally fluoridated soft water, artificially fluoridated soft water, naturally fluoridated hard water and artificially fluoridated hard water) and compared with reference water, using a measure of relative bioavailability. All four drinking waters studied contained fluoride at approximately 1 ppm (1 mg F L 1). Results of this research are said by the authors to indicate no statistically significant difference between artificially fluoridated and naturally fluoridated water, or between hard and soft water, for any of the parameters investigated following water ingestion by healthy young adults. Within the limits of the study methodology, this work found no evidence for any differences between the absorption of fluoride ingested in artificially fluoridated drinking water and in drinking water in which the fluoride is present naturally, or between the absorption of fluoride from hard and soft waters, at fluoride concentrations close to 1 ppm. These results are important because they demonstrate that results of studies on populations receiving naturally fluoridated drinking water can continue to be applied to the assessment of effects in populations exposed to similar fluoride levels artificially. The most important markers of exposure are measured fluoride levels in plasma, urine and bone. The fluoride content of tooth tissues reflects the biologically available fluoride at the time of tooth formation and, for the majority of the enamel, once formed, fluoride levels remain constant [19]. Bone provides a measure of cumulative exposure and is probably the best guide to long-term uptake, but it can be difficult to obtain and is not practical for largescale studies or routine clinical situations. Plasma levels give the best practicable indication of recent fluoride intake; fluoride levels in saliva reflect those in plasma [16,20]. Although a useful marker of absorbed dose, urinary excretion of fluoride is of somewhat limited value for estimating fluoride ingestion because of variation in the proportion of ingested fluoride that is retained (rather than excreted), which depends mainly on the level of fluoride intake. Also, fluoride excretion is dependent on the pH of the urine and there appear to be age-related differences in the proportion of ingested fluoride that is excreted in the urine. Nevertheless, urinary fluoride measurements can be useful markers of recent exposure, especially if 24 h samples are used, and are valuable in comparative studies [12]. 3. Health impacts

1451

thought to be three main ways by which fluoride modifies the process of tooth decay: improvement of the chemical structure of the enamel during development, making it more resistant to acid attack; the encouragement of remineralisation with an improved quality of enamel crystals and reduction of the ability of plaque bacteria to produce acid [3]. The York Review [10] found evidence that water fluoridation is effective in reducing dental caries, and that the reduction in dental caries experience is greater in those areas with higher levels of dental caries prior to water fluoridation. The change in the prevalence of dental caries was an estimated 15% increase in the proportion of subjects with no dental caries and a decrease of 2.2 in the mean number of decayed, missing or filled teeth (dmft/DMFT).5 However, due to the variability between reported studies it was not possible to be precise about the size of these effects, and many studies had failed to take sufficient account of confounding factors. In considering the effects on dental caries, the York Review looked only at those studies presenting baseline and follow-up data for both fluoridated and non-fluoridated communities and it considered only two outcome measures: differences in percentages of caries-free children and absolute differences in caries experience (i.e. differences in dmft/DMFT). Other reported studies, such as cross-sectional studies6 conducted in the UK between 1980 and 1990, have shown reductions in dental caries due to fluoride of the same order of magnitude as those reported in the York Review [21–27]. These studies also reported additional effects, such as reductions in the prevalence of toothache and of dental treatment needing general anaesthesia. 3.1.1. Effects of social class As with many diseases in the UK, dental caries (especially in the primary dentition of young children) is more prevalent in the more deprived social groups than in more affluent social groups [28,29]. Moreover, high prevalencies of toothache, abscesses and dental extractions needing general anaesthesia are associated with the high caries experience of children in deprived social groups in the UK [21,30,31]. The two principle factors influencing dental caries are diet and the use of fluoridated dental care products (especially toothpaste). Although the reduction in sugar consumption in UK children, since the 1960s and the introduction of fluoride toothpaste in the 1970s led to substantial reductions in dental caries [32], these reductions were not uniform and led to widening social inequalities in children’s dental health. Diets of more socially deprived children are more caries-conducive than diets of more affluent children, and more affluent children brush

3.1. Dental caries Although the relationship between fluoride and tooth decay is complex, and not yet fully understood, there are currently 4 Available at: http://www.ncl.ac.uk/dental/assets/docs/fluoride-bioavailability%20of%20fluoride%20in%20drinking%20water%20REPORT.pdf.

5 dmft: mean number of decayed, missing or filled teeth in the deciduous dentition (first teeth). DMFT: mean number of decayed, missing or filled teeth in the permanent dentition. 6 Cross-sectional studies are epidemiological studies that measure the prevalence of health outcomes of determinants of health (or both) in a population at a point in time or over a short period of time. Such studies are often used to explore the causes of disease.

1452

P.T.C. Harrison / Journal of Fluorine Chemistry 126 (2005) 1448–1456

their teeth with a fluoride toothpaste more often than do more socially deprived children [31]. 3.2. Dental fluorosis Dental fluorosis is a form of developmental defect of tooth enamel. Histologically it presents as a hypocalcification, while clinically it ranges from barely visible white striations on the teeth through to gross defects and staining of the enamel. There are around 90 different causes of enamel defects, of which three or four causes are common. Minor forms of dental fluorosis are not aesthetically troublesome and may even enhance the appearance of dental enamel [33]. The York Review identified 88 studies (mainly crosssectional) investigating dental fluorosis, from 30 countries, which suggested a prevalence (all levels of severity) of 48% in fluoridated areas and 15% in non-fluoridated areas. Limiting consideration to aesthetically important levels of severity, the York Review reported the prevalence of fluorosis to be 12.5% in fluoridated areas and 6.3% in non-fluoridated areas. For any given fluoride concentration in water the prevalence of aesthetically important dental fluorosis was higher in naturally fluoridated areas than in artificially fluoridated areas. A sensitivity analysis excluding data points above 1.5 ppm fluoride found prevalences for all levels of severity of 46 and 18% and for aesthetically important dental fluorosis of 10 and 6% in fluoridated and non-fluoridated areas, respectively. However, as pointed out by the MRC Working Group [12], the York Review included studies in countries with hotter climates than the UK; in hot climates, water intake is typically higher than in the UK and the risk of fluorosis correspondingly greater for any given water fluoride concentration [34]. Higher incidences of dental fluorosis have been reported in children in the USA compared with the UK. However, studies have suggested that 7–36% of children living in fluoridated US communities may also be receiving fluoride supplements inappropriately prescribed by their physician or paediatrician [35], which could contribute to the higher prevalence values reported, since discretionary fluoride products are an important aetiological factor for dental fluorosis. In addition, low fluoride toothpastes have not been marketed in the USA. 3.3. Bone health The York Review [10] included 29 studies on the relation of fluoride in water to bone health. These covered fractures at various anatomical sites, slipped epiphysis and otosclerosis. Eighteen of the investigations provided data on hip fracture. The validity and evidence value of the studies was generally assessed as low. The estimated relative risk (obtained from 20 studies) for fracture associated with a fluoride concentration of 1 ppm in water ranged either side of the null value, with a pooled estimate from a univariate meta-regression of 1.00 (95% CI 0.94–1.06). Two studies of otosclerosis both suggested a beneficial effect of fluoridation, and in a single investigation of slipped epiphysis, fluoride in water was associated with an

increased risk in boys and a reduced risk in girls, neither of which was statistically significant. An effect of fluoridation on the risk of fracture, adverse or beneficial, is biologically plausible. Fluoridation of water can increase normal dietary intake of the mineral by some 50%, and about half of the fluoride ingested is taken up by bone. Within the bone, fluoride ions can replace hydroxyl ions in the hydroxyapatite lattice, with possible implications for its mechanical properties. In addition, elevation of the fluoride concentration in plasma directly increases osteoblastic differentiation and activity. In theory, a number of other bone disorders could also be affected by these mechanisms. For example, alterations in the hydroxyapatite lattice might influence the development of otosclerosis [12]. Many of the epidemiological studies on fluoride and bone health have assessed risk only in relation to current or recent exposure to fluoridated water. However, given the possible mechanisms for an effect on bone, a more relevant metric is likely to be some index of cumulative exposure. This was explored in a recent MRC case-control study of hip fracture, which found no elevation of risk with exposures to higher fluoride concentrations over a lifetime [36]. A possible limitation of that study, however, was that the exposure to fluoride was almost all from natural sources in water that also contained high concentrations of calcium, and it is known (see above) that calcium might reduce the bioavailability of fluoride from the gastrointestinal tract. 3.4. Cancer The possibility that fluoridation might increase the risk of developing cancer was raised by a series of reports of experiments in mice [37,38] and by a report in 1975 purporting to show higher overall cancer mortality rates among the 10 largest US cities that practised water fluoridation than amongst the 10 largest US cities that did not [39,40]. Neither the results of these early animal experiments nor the report of Burk and Yiamouyiannis were accepted by subsequent expert reviews (e.g. by IARC [41] and Knox [42]), but the important public health implications of the question have stimulated many further investigations, including a lifetime rodent study [43]. Particular attention has been given to bone cancer, especially osteosarcoma, because ingested fluoride is concentrated in the bones. Some attention has also been given to cancers of the stomach, kidney and thyroid [12]. Overall, the current evidence does not support the hypothesis that exposure to artificially fluoridated water causes an increase in the risk for cancer in humans. It is too early to see whether there might be an effect after very long exposure, but the results available rule out more than a very small effect of artificial fluoridation on cancer risk for up to about 35 years of exposure [12]. Furthermore, studies of cancer rates in relation to variations in naturally occurring fluoride levels provide information on lifetime exposure and the absence of any detectable adverse effects of fluoride in these studies gives a high level of reassurance concerning safety [42]. The features and consequences characteristic of high fluoride ingestion in humans and other animals have not

P.T.C. Harrison / Journal of Fluorine Chemistry 126 (2005) 1448–1456

included the occurrence of cancer. Most agents that cause cancer directly do so because they are genotoxic, although some (nongenotoxic) agents can cause or promote cancer by other mechanisms, for example, by stimulating cell division [12]. For fluoride, in vitro genotoxicity data are mostly for doses much higher than those to which humans are exposed. Even at these high doses, genotoxic effects are not always observed [44], and fluoride is consistently negative in the Ames test [45]. Some in vivo studies have indicated that fluoride can in some circumstances induce mutations and chromosome aberrations in rodent and human cells, and in particular it has been shown that fluoride can have a mitogenic effect on osteoblasts [43], which could provide a mechanism by which fluoride might increase the risk for osteosarcoma (primary bone cancer). Overall, the evidence available has not established that fluoride is genotoxic in humans, and most of the studies suggest that it is not, but the possibility of some genotoxic effect cannot be excluded [44,45]. The specific evidence for a possible mitogenic effect of fluoride on osteoblasts led the MRC Working Group to recommend that any future studies on the incidence and cause of this rare bone cancer should included an assessment of fluoride exposure. They also suggested that it would be prudent to undertake an updated analysis of UK data on fluoridation and cancer rates generally [12]. 3.5. Other health effects Fluoride exposure has been postulated to cause a number of health effects other than those described above. Many of these, although plausible, have not been substantiated. Some of the more important are outlined below. 3.5.1. Immunological effects Information regarding the allergic potential of fluoride in drinking water is sparse. A paper by Spittle [46] concluded that some individuals exhibit an allergic/hypersensitivity reaction to fluoride, but reviews by NRC [44], NHMRC [47] and Chalacombe [48] all concluded that the studies undertaken do not support claims that fluoride is allergenic. They considered the weight of evidence to show that fluoride is unlikely to produce hypersensitivity or other immunological effects. There is no information on the immunotoxicity of fluoride. 3.5.2. Effects on reproduction Adverse effects of fluoride intake on reproductive performance, such as reduced lactation, have been demonstrated in many species. However, these studies have used dietary concentrations very much higher than those in the fluoridated drinking water of humans [44]. Fluoride has also been implicated in a number of adverse outcomes relating to fertility and pregnancy, but there is insufficient evidence to establish a link between decreased fertility and fluoride exposure [14]. The York Review found no evidence of reproductive toxicity in humans [10]. A multigenerational study of sodium fluoride in rats, at fluoride levels in drinking water of up to 250 ppm, found no impacts on reproduction, and mating fertility and survival

1453

indices were not affected [49]. Parallel studies using the same exposure regimen revealed no evidence for effects on testis structure, spermatogenesis or endocrine function in male rats [50,51], nor on numbers of corpora lutea, implants and viable foetuses in females [49]. 3.5.3. Birth defects Fluoride crosses the placenta and is incorporated in the tissues of the developing conceptus. Studies in areas of India and Africa that have high levels of naturally fluoridated water have not shown an increase in birth defects [45]. Erickson et al. [52] found an association between drinking fluoridated water and congenital malformations in one set of data, but not in another. A study in Atlanta, Georgia, using the birth defects registry, found no association between birth defects and fluoridation of community water supplies [45]. In 1957, an investigator linked an excess of Down’s syndrome to fluoridation. However, later studies by other investigators provided strong evidence against this suggestion [10,45]. The York Review [10] reported six studies that examined whether there is an association between Down’s syndrome and drinking water fluoride level [55].7 All of the studies were of poor quality according to the review criteria. Four of the studies showed no significant association. Two studies [53,54] found a significant ( p < 0.05) positive association, i.e. increased Down’s syndrome incidence with increased water fluoride level. However, it was noted that these two positive studies had methodological limitations; for example, they did not control appropriately for various possible confounding effects, including maternal age. Thus, the evidence for an association between water fluoride level and the incidence of Down’s syndrome is inconclusive, a conclusion reiterated by Whiting et al. [55]. If fluoride reaches the developing foetus and is incorporated into its tissues, it could plausibly be teratogenic. The DHSS review [45] concluded that experimental animal data do not provide any additional evidence for an association between fluoride in drinking water and birth defects; the other major reviews [14,44,47] provide no comment on this issue. A recent multi-generation developmental toxicity study on rats given up to 250 ppm fluoride in drinking water [49] showed no effects on foetal morphological development, although ossification of the hyoid bone in F2 foetuses was significantly reduced at the top dose level. Human and experimental animal data suggest that drinking even high levels of fluoride in water does not cause birth defects, though there may be adverse consequences for bone ossification at very high exposure levels [10]. 3.5.4. Renal effects The kidney is a potential site of acute fluoride toxicity because of its exposure to relatively high fluoride concentrations [44]. It has been established from human studies that the kidney removes fluoride from the blood more efficiently than it 7 The York team has subsequently published a paper specifically on Down Syndrome and water fluoride levels (Whiting et al. [55]).

1454

P.T.C. Harrison / Journal of Fluorine Chemistry 126 (2005) 1448–1456

removes other halides. In addition, renal clearance of fluoride decreases in individuals with renal insufficiency or diabetes mellitus. However, several large community-based epidemiological studies found no increased renal disease associated with long-term exposure to drinking water with fluoride concentrations of up to 8 mg L 1 [44,45]. 3.5.5. Gastrointestinal tract With the exception of monofluorophosphate, high concentrations of fluoride-releasing compounds form hydrogen fluoride on mixing with hydrochloric acid in the stomach. Hydrogen fluoride can be irritating to the gastric mucosa, resulting in dose-dependent adverse effects. The data for human effects at low exposure are limited, but the indication is that gastrointestinal effects are not a problem at optimal drinking water fluoride concentrations [44,45]. A study by Susheela et al. [56] assessed the prevalence and severity of gastrointestinal disturbances (and other non-skeletal manifestations) in an area of endemic skeletal and dental fluorosis in India. The highest prevalence (52.4%) of non-ulcer dyspeptic symptoms was found among 288 individuals (69 families) living in a village where the (natural) mean fluoride concentration in the 36 separate water sources was 3.2 ppm (range 0.25–8.0 ppm). Eleven of these water sources were defined by the authors as ‘safe’ (i.e. with fluoride levels of 1.0 ppm or less). The authors noted that in patients who reverted to ‘safe’ water, dyspeptic symptoms and complaints disappeared within 2–3 weeks. Other research by Susheela et al. [57] revealed that the long-term ingestion of fluoride by ten patients on sodium fluoride therapy (30 mg per day) for otosclerosis was associated with non-ulcer dyspeptic symptoms in eight of the patients. The effects of fluoride on the gastric mucosa have been described in detail by Whitford [16]. Gastric irritation, by release of hydrogen fluoride in the stomach at high doses of fluoride intake, is plausible. However, it is unlikely that sufficient hydrogen fluoride will be released from the low concentrations of fluoride in drinking water in the UK to cause irritation in healthy individuals. It is possible that individuals who have an existing stomach disorder may be susceptible to irritation following ingestion of fluoridated water, but there is no published evidence for this [12]. 3.5.6. Intelligence Two Chinese studies [58,59] have found a positive association between high levels of fluoride in drinking water and reduced children’s intelligence/IQ. Confounding factors were dismissed, but their possible influence on the results of the study was not adequately explained by the authors. At lower fluoride concentrations (e.g. 0.91 ppm), which are more comparable to the levels in fluoridated water in the UK, a reduction in children’s IQ was not observed. There is a possible link here with lead toxicity and the impact of fluoride on lead bioavailability (see below). 3.5.7. Thyroid (goitre) The York Review listed three studies in which goitre was the outcome of interest. Two of these studies [60,61] found no

significant association with water fluoride level. The third [62] found a significant positive association between combined high fluoride/low iodine levels and goitre. However, this study examined combined fluoride/iodine uptakes, and has not been published in a peer-reviewed journal. 3.5.8. Miscellaneous effects Several other health outcomes have been postulated as being connected with elevated fluoride intake, including effects on the pineal gland, senile dementia, age at menarche, anaemia during pregnancy, Sudden Infant Death Syndrome (SIDS) and primary degenerative dementia. Available information on these outcomes is limited and inconclusive [12]. 3.6. Indirect effects of adding fluoride to water In addition to any direct impact on health resulting from increased uptake of fluoride by the body, it is possible that fluoridation of water supplies could influence health through other mechanisms; for example:  toxicity from other substances added to water as part of the fluoridation process;  an effect of higher fluoride in water on dietary exposure to toxic metals (e.g. through leaching of metals from pipework and cooking pans);  an effect of fluoride in drinking water on the uptake/ bioavailability or toxicity of metals in the gut. 3.7. Substances added during the fluoridation process The UK’s Water (Fluoridation) Act 1985 allows hexafluorosilicic acid (H2SiF6) and disodium hexafluorosilicate (Na2SiF6) to be used to increase the fluoride content of water. The published Code of Practice on Technical Aspects of Fluoridation of Water Supplies [63] gives specifications for these substances and states that ‘the product. . . must not contain any mineral or organic substances capable of impairing the health of that drinking water correctly treated with the product’. For H2SiF6, limits are given for a number of possible impurities, including for iron, heavy metals, sulfate, phosphate, and chloride. The specification for Na2SiF6 powder requires a minimum of 98% m/m of the pure chemical, and gives maximum limits for impurities, including heavy metals (as lead) and iron. No other substances are allowed to be used in the fluoridation process, other than an anti-caking agent (the identity of which must be disclosed) in the case of Na2SiF6. Synthetic detergents are not permitted. Thus there is no likelihood, in normal operation, for any fluoridation plants to introduce other compounds into the drinking water supply (other than approved anti-caking agents and any impurities present in the fluoridation chemicals). It has been suggested that arsenic is introduced into drinking water through the fluoridation process because this element is present as an impurity in fluoride compounds. However, because of the dilution factor, the contribution of arsenic from this source would be extremely small, and in any case there is a standard

P.T.C. Harrison / Journal of Fluorine Chemistry 126 (2005) 1448–1456

for the total allowable arsenic level in drinking water [12]. The review by Jackson et al. [18] similarly concluded that the traces of impurities added as a result of fluoridation are very small and would have no measurable impact on the toxicity of drinking water. 3.8. Dietary exposure to metals Fluoride in water at normal levels can increase slightly the amount of leaching of aluminium from cooking utensils [64]. High concentrations of fluoride can also result in leaching of copper from pipework, but this is not likely to be significant at a fluoride level of 1 ppm [18]. Overall, such effects are considered to be of minimal health significance in normal circumstances [12]. 3.9. Effects on bioavailability or toxicity of toxic metals A number of studies and theoretical considerations have indicated the possibility that fluoride may affect the uptake, bioavailability and/or toxic effects of aluminium and lead [65,66]. According to Jackson et al. [18], the presence of fluoride at 1 ppm will have practically no effect on the chemical reactions and uptake of lead (or of iron or copper), but fluoride can form strong bonds with aluminium in slightly acidic water; moreover, complexities associated with speciation, ionic interactions, etc., yield uncertainties in a number of aspects, and the MRC Working Group concluded that this area merits continued surveillance [12]. 4. Conclusions Fluoride in water contributes significantly to the total exposure of an individual to this element, but it is not the only source of exposure, especially since the introduction of fluoride toothpastes. There is evidence that water fluoridation reduces caries experience in children but that dental fluorosis (mottling of teeth) is a possible unwanted consequence. Because of the wide use of toothpastes and other dental health care products containing fluoride, and the potential for fluoride exposure from a number of other sources, it is especially important to understand better the total exposure that individuals are experiencing. There is almost universal agreement that tooth decay in children is related to social class, and much of the research conducted to date indicates that water fluoridation reduces dental caries inequalities between high and low social groups. There are a number of possible health outcomes (other than dental health) related to water fluoridation. The possibility of an effect on the risk of hip fracture is the most important in public health terms. The available evidence on this suggests no effect, but cannot rule out the possibility of a small percentage change (either an increase or a decrease) in fractures. Research results currently available do not allow a useful estimate to be made of the impact of fluoridation on other bone disorders. However, the few studies that have been carried out do not suggest a problem.

1455

Available evidence suggests no link between water fluoridation and either cancer in general or any specific cancer type (including osteosarcoma, primary bone cancer); nonetheless, the MRC Working Group recommended an updated analysis of UK data on fluoridation and cancer rates and suggested that if new studies are undertaken on the incidence and causes of osteosarcoma then fluoride exposure should be assessed together with the other possible risk factors [12]. Additional health outcomes suggested by some to be associated with fluoride ingestion include effects on the immune system, reproductive and developmental (birth) defects and effects on the kidney and gastrointestinal tract. Other concerns are related to the chemicals that are added during the fluoridation process, and to indirect effects, such as increased leaching of lead from pipes and aluminium from cooking utensils and altered uptake or toxicity of these substances. The evidence for any significant health effects of this type is considered to be weak, although the MRC Working Group recommended that the area be kept under review [12]. Acknowledgement I wish to acknowledge the indirect contributions made to this paper by members of the MRC Working Group on Water Fluoridation, which I had the honour and pleasure to chair. References [1] M. Hamilton, J. Environ. Health 54 (1992) 27–32. [2] J.J. Murray, A.J. Rugg-Gunn, G.N. Jenkins, Fluorides in Caries Prevention, third ed., Butterworth-Heinemann, Oxford, 1991. [3] British Fluoridation Society, One in a Million: the Facts about Water Fluoridation, British Fluoridation Society, London, 2004. [4] US Department of Health and Human Services, Surgeon General Statement on Water Fluoridation (28 July), Office of the Surgeon General, DHSS, Rockville, MD, 2004 (available at: http://www.cdc.gov/ OralHealth/pdfs/SGstatement.pdf). [5] Ministry of Health, Scottish Office, Ministry of Housing and Local Government, The Conduct of the Fluoridation Studies and the Results Achieved after Five Years, Reports on Public Health and Medical Subjects No.105, Her Majesty’s Stationary Office (HMSO), London, 1962. [6] Department of Health and Social Security, Scottish Office, Welsh Office, Ministry of Housing and Local Government, The Fluoridation Studies in the United Kingdom and the Results Achieved after Eleven Years, Reports on Public Health and Medical Subjects No.122, HMSO, London, 1969. [7] House of Commons Official Report, Parliamentary Debates, Wednesday 6 May, Hansard, 311, col 697, 1998. [8] H.D. Foster, J. Orthomol. Med. 8 (1993) 149–153. [9] MRC Working Group, Internal Report 94/BO95 on Fluoridation of Drinking Water – Link with Osteoporosis, 1994. [10] NHS CRD, A Systematic Review of Public Water Fluoridation, CRD Report No. 18, NHS Centre for Review and Dissemination, University of York, York, UK, 2000 (available at: http://www.york.ac.uk/inst/crd/ fluorid.htm). [11] Department of Health, Saving Lives: Our Healthier Nation, Cm 4386, Stationery Office, London, 1999 (available at: http://www.officialdocuments.co.uk/document/cm43/4386/4386.htm). [12] Medical Research Council, Working Group Report, Water Fluoridation and Health, Medical Research Council, London, 2002. [13] R.L. Akenhurst, D.J. Sanderson, Br. J. Med. Econ. 7 (2004) 43–54. [14] NHMRC, Review of Water Fluoridation and Fluoride Intake from Discretionary Fluoride Supplements, National Health and Medical Research

1456

[15] [16] [17] [18] [19]

[20] [21] [22] [23] [24] [25] [26] [27] [28] [29]

[30] [31]

[32] [33] [34] [35]

[36] [37] [38] [39]

P.T.C. Harrison / Journal of Fluorine Chemistry 126 (2005) 1448–1456 Council, Melbourne, 1999 (available at: http://www.health.gov.au/hfs/ nhmrc/advice/pdf/fluoride.pdf). J.J. Warren, S.M. Levy, Dent. Clin. N. Am. 43 (1999) 695–711. G.M. Whitford, Monographs in Oral Science, vol. 16, The Metabolism and Toxicity of Fluoride, second revised ed., Karger, Basel, 1996. H.-D. Cremer, W. Buttner, Absorption of fluorides, in: Fluorides and Human Health, World Health Organization, Geneva, 1970. P. Jackson, P. Harvey, W. Young, Chemistry and Bioavailability Aspects of Fluoride in Drinking Water, WRc-NSF, Marlow, 2002. WHO, Fluorides and Oral Health: Report of a WHO Expert Committee on Oral Health Status and Fluoride Use, Technical Report Series 846, World Health Organization, Geneva, 1994. G.N. Jenkins, The Physiology and Biochemistry of the Mouth, Oxford, Blackwell, Oxford, 1978. A.J. Rugg-Gunn, C.L. Carmichael, A.D. French, J.A. Furness, Br. Dent. J. 142 (1977) 359–402. A.D. French, C.L. Carmichael, A.J. Rugg-Gunn, J.A. Furness, Br. Dent. J. 156 (1984) 54–57. J.T. Duxbury, M.A. Lennon, C.M. Mitropoulos, H.V. Worthington, Br. Dent. J. 162 (1987) 457–458. C.M. Mitropoulos, M.A. Lennon, J.W. Langford, D.J. Robinson, Br. Dent. J. 164 (1988) 349–350. A.J. Rugg-Gunn, C.L. Carmichael, R.S. Ferrell, Br. Dent. J. 165 (1988) 359–364. C.L. Carmichael, A.J. Rugg-Gunn, R.S. Ferrell, Br. Dent. J. 167 (1989) 57–61. S. Seaman, F.D. Thomas, W.A. Walker, Commun. Dent. Health 6 (1989) 215–221. D. Acheson, Independent Inquiry into Inequalities in Health, HMSO, London, 1998. J.F. Beal, Social factors and preventive dentistry, in: J.J. Murray (Ed.), The Prevention of Oral Diseases, Oxford University Press, Oxford, 1996, pp. 217–233. M. O’Brien, Children’s Dental Health in the United Kingdom 1993, HMSO, London, 1994. K. Hinds, J.R. Gregory, National Diet and Nutrition Survey: Children Aged 1.5 to 4.5 Years, vol. 2, Report of the Dental Survey, Stationery Office, London, 1995. J.E. Todd, T. Dodd, Children’s Dental Health in the United Kingdom 1983, HMSO, London, 1985. G.M. Hawley, R.P. Ellwood, R.M. Davies, Commun. Dent. Health 13 (1996) 189–192. J.J. Murray, Appropriate Use of Fluorides for Human Health, World Health Organization, Geneva, 1986. US Department of Health and Human Services Centers for Disease Control and Prevention, Morbidity and Mortality Weekly Report, vol. 50, Number RR-14, Recommendations for Using Fluoride to Prevent and Control Dental Caries in the United States, US Department of Health and Social Services, Atlanta, GA, 2001. S. Hillier, C. Cooper, S. Kellingray, G. Russell, H. Hughes, D. Coggon, Lancet 355 (2000) 265–269. A. Taylor, Dent. Digest 60 (1954) 170–172. A. Taylor, N.C. Taylor, Proc. Soc. Exper. Biol. Med. 119 (1965) 252–255. D. Burk, J. Yiamouyiannis, Letter July 8 to Hon. James J. Delaney, Congressional Record, Proceedings and Debates of the 94th Congress, First Session, House of Representatives, July 21, 1975, pp. 23729–23732.

[40] J. Yiamouyiannis, D. Burk, Fluoride 10 (1977) 102–123. [41] IARC Monographs on Evaluation of Carcinogenic Risks of Chemicals to Humans, vol. 27, Some Aromatic Amines, Anthraquinones and Nitroso Compounds, and Inorganic Fluorides Used in Drinking Water and Dental Preparations, International Agency for Research on Cancer, Lyon, 1982, pp. 279–303. [42] E.G. Knox, Fluoridation of Water and Cancer: A Review of the Epidemiological Evidence, Report of the Working Party on the Fluoridation of Water and Cancer, HMSO, London, 1985. [43] J.R. Bucher, M.R. Hejtmancik, J.D. Toft II., R.L. Persing, S.L. Eustis, J.K. Haseman, Int. J. Cancer 48 (1991) 733–737. [44] NRC, Health Effects of Ingested Fluoride, National Academy Press, Washington DC, 1993. [45] DHHS, Review of Fluoride—Benefits and Risks, Department of Human Health and Services, Washington DC, 1991. [46] B. Spittle, Fluoride 26 (1993) 267–273. [47] NHMRC, The Effectiveness of Water Fluoridation, Report D6, National Health and Medical Research Council, Canberra, 1991. [48] S.J. Challacombe, Commun. Dent. Health 13 (Suppl. 2) (1996) 69–71. [49] T.F.X. Collins, R.L. Sprando, T.N. Black, M.E. Shackelford, N. Olejnik, M.J. Ames, J.I. Rorie, D.I. Ruggles, Food Chem. Toxicol. 39 (2001) 867– 876. [50] R.L. Sprando, T.F. Collins, T.N. Black, J. Rorie, M.J. Ames, M. O’Donnell, Food Chem. Toxicol. 35 (1997) 881–890. [51] R.L. Sprando, T.F. Collins, T.N. Black, N. Olejnik, J. Rorie, Food Chem. Toxicol. 36 (1998) 1117–1124. [52] J.D. Erickson, G.P. Oakley, J.W. Flynt, S. Hay, J. Am. Dental Assoc. 76 (1976) 981. [53] I. I. Rapaport, Bull. Acad. Nat. Med. 140 (1957) 529–531. [54] I. Rapaport, Rev. Stomatol. Chir. Maxillofac. 46 (1963) 207–218. [55] P. Whiting, M. McDonagh, J. Kleijnen, BMC Public Health, vol. 1, 2001, p. 6 (available at: www.biomedcentral.com/1471-2458/1/6). [56] A.K. Susheela, A. Kumar, M. Bhatnagar, R. Bahadur, Fluoride 26 (1993) 97–104. [57] A.K. Susheela, T.K. Das, I.P. Gupta, Fluoride 25 (1992) 5–22. [58] Y. Lu, Z.R. Sun, L.N. Wu, X. Wang, W. Lu, S.S. Liu, Fluoride 33 (2000) 74–78. [59] L.B. Zhao, G.H. Liang, D.N. Zhang, X.R. Wu, Fluoride 29 (1996) 190– 192. [60] I. Gedalia, N. Brand, Arch. Int. Pharmacodyn. 142 (1963) 312–315. [61] P. Jooste, M. Weight, J. Kriek, A. Louw, Eur. J. Clin. Nutr. 53 (1999) 8–12. [62] F.-F. Lin, H.-X. Zhao, J.-Y. Jian, The Relationship of a Low Iiodine and High Fluoride Environment to Subclinical Cretinism in Xinjiang, Xinjiang Institute for Endemic Disease Control and Research, Office of Leading Group for Endemic Disease Control of Hetian Prefectural Committee of the Communist Party of China and County Health and Endemic Prevention Station, Yutian, Xinjiang, 1991 (unpublished report submitted through NHS CRD website). [63] Department of the Environment, Code of Practice on Technical Aspects of Fluoridation of Water Supplies, HMSO, London, 1987. [64] G.H. Moody, J.C. Southam, S.A. Buchan, J.G. Farmer, Br. Dent. J. 169 (1990) 47–50. [65] J.A. Varner, K.F. Jensen, W. Horvath, R.L. Isaacson, Brain Res. 784 (1998) 284–298. [66] R.D. Masters, M.J. Coplan, B.T. Hone, J.E. Dykes, Neurotoxicology 21 (2000) 1091–1100.