Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
2 DENTAL FLUOROSIS THE FLUOIUDE CONTENT OF n ETA A,! humans ingest fluoride to some extent, and fluoride's affinity for calcified tissues makes it a normal constituent of dental tissues. Up to 4 times as much fluoride is contained in dentin, the bone-like material that constitutes the bulk of a tooth, as in enamel, the visible outer layer. Fluoride is not evenly distributed in enamel and is concentrated primarily on the outer enamel surface. Most fluoride is incorporated into the crystalline lattice of enamel before tooth eruption, but more is incor- porated into the enamel crystals immediately after eruption, while the enamel is still maturing. The capacity for sound, newly erupted] teeth to absorb fluoride into the enamel's crystalline lattice, however, rapidly diminishes as the enamel matures (Weatherell et al., 1977~. As a result, the amount of fluoride in dental enamel does not increase with age to nearly the same extent as in bone. Enamel that has undergone clemineral- ization, the first histological effect of dental caries, and subsequent remineralization contains higher concentrations of fluoride than does sound enamel (Silverstone, 1977~. The concentration of fluoride in sound, mature tooth enamel, at a depth of approximately 2 micrometers (#m), averages 1,700 ppm in people residing in areas with low concentrations of fluoride in drinking 19
20 Health Effects of Ingested Fluoride water (i.e., fluoride at 0.l mg/L or less) and 2,200-3,200 ppm in areas with concentrations of fluoride at approximately I.0 mg/L. Enamel fluoride concentrations of this order have no adverse impact on oral health; indeed, the presence of fluoride in the crystalline lattice of dental enamel can only increase the enamel's resistance to dissolution in decay- causing acids. When drinking water contains naturally occurring fluoride at 5-7 mg/1= (much higher than recommended), enamel fluoride concen- trations have been measured at 4,800 ppm (Aasenden, 1974~. No data could be found on enamel fluoride concentrations in people residing in areas where drinking water contains I-5 mg/~. People from areas with high concentrations of naturally occurring fluoride in drinking water (i.e., 5-7 mg/~) usually exhibit severe dental fluorosis, and their enamel can become brittle enough to fracture at incisal edges and cusp tips. Caries might begin in the broken enamel, and, even if it does not, teeth in this condition often require treatment to restore function. ~. ~ n^^ ~ ~ A ~1 n7AN FLUORIDE'S ACTION IN PnEvENTING DENTAL CARIES Various forms of fluoride have been used in American dentistry over the past 40 years to prevent tooth decay. The results have been highly successful, and the prevalence and severity of dental caries in the 199Os are substantially reduced from the levels seen in the 19SOs. Fluoride's action in preventing caries comes from both pre-eruptive and post-eruptive mechanisms (Dawes, 1989~. In some studies of water fluoridation, the greatest reductions In caries were seen in children who were born as fluoridation began or thereafter, evidence that supports a pre-eruptive effect (Grainger and Coburn, 1955; Forrest and James, 1965; Horowitz and Heifetz, 1967; Katz and Muhier, 1968; Chandra et al., 1980~. However, evidence of caries-preventive effects in teeth already erupted or in the process of erupting when fluoridation began is also long-standing (Klein, 1945, 1946; Arnold et am 1953; Ast and Chase, 1953; Backer Dirks et al., 1961; Russell and Hamilton, 1961; Backer Dirks, 1967) and has been confirmed more recently (Hardwick et al., 1982~. The relative importance of pre-eruptive and post-eruptive effects continues to be researched and debated (Groeneveld et al., 1990; 7 ~
Dental Fluorosis 21 Horowitz, 1990; Ttlylstrup, 1990), although the mechanisms of the effects are well-def~ned. Pre-eruptively, ingested fluoride is incorporated into the developing enamel hydroxyapatite crystal, where it has the effect of reducing enamel solubility (Beltran and Burt, 1988~. Pre-eruptive fluoride might prevent caries more effectively in pit-and-fissure lesions than in smooth-surface caries (Groeneveld et al., 1990~. Post-eruptively, the frequent infusion of low-concentration fluoride into the oral cavity, such as drinking fluoridated water or regular brushing with a fluoride toothpaste, enhances remineralization (i.e., rebuilding of the enamel matrix) when demineralization has occurred! in the early stages of the curious process (Koulourides, 1990~. Some of the fluoride that comes into the mouth from water, food, or toothpaste concentrates in dental plaque (Singer et al., 1970), where most of it is held in bounct rather than free ionic form. The bound fluoride can be releaser! in response to lowered plaque pH that demineralizes enamel (Tatevossian, 1990), ant] fluoride is taken up more readily by demineralized enamel than by sound enamel (White and NancolIas, 19901. The availability of plaque fluoride to respond to acid challenge leads to gradual establish- ment of well-crystallized and more acid-resistant apatite in the enamel surface during the frequent demineralization-remineralization cycles (Ericsson, 1977; TtlyIstrup et al., 1979; Kidd et al., 1980; Chow, 1990; ThyIstrup, 1990~. Plaque fluoride also inhibits glycolysis, the process in which sugar is metabolized by bacteria to produce acid. Plaque fluoride can retard the production of extracellular polysaccharide by cariogenic bacteria, a production that is necessary for plaque to adhere to smooth enamel surfaces and, hence, for caries to occur (Hamilton, 1990~. In assessing the relative impact of pre-eruptive ant! post-eruptive effects of fluoride, it became evident by the mid-1970s that a high con- centration of enamel fluoride could not by itself explain the extensive reductions in caries that fluoride produced (Levine, 1976~. As clescribed earlier, enamel fluoride concentrations at a depth of 2 Am average I,700- 4,XOO ppm, depending on age and fluoride exposure. The theoretical concentration of fluoride in pure fluorapatite that would reduce its acid! solubility to the extent necessary to explain all reductions in caries is approximately 3S,000 ppm (Wefe} et al., 1986~. It has also been shown
22 Health Effects of Ingested Fluoride that high concentrations of enamel fluoride do not necessarily mean that caries will not occur (Arenas and Christoffersen, 1990~. The evidence collectively demonstrates that fluoride prevents caries by post-eruptive actions as well as by pre-eruptive incorporation of hydroxyapatite crystals into the enamel. In addition, high-concentration fluoride at approximately 12,000 ppm, as used in professionally applied gels, might have a specific bactericidal action on cariogenic bacteria in plaque (Bowden, 1990~. Those gels also leave a temporary layer of calcium fluoride on the enamel surface, which is available for release when the pH at the enamel surface is lowered ~eGeros, 1990~. The cariogenic bacterium Streptococcus mutans has been shown to become less acidogenic through adaptation to an environ- ment where it is regularly exposed to low concentrations of fluoride in drinking water or to higher concentrations in toothpastes and mouth rinses (Rosen et al., 1978; Bowden, 1990; Marquis, 1990~. It is plaus- ible, though not confirmed, that this ecological adaptation reduces the cariogenicity of S. mutans in humans (van Loveren, 1990~. HISTOPATHOLOGY OF DENTAL FLUOROSIS Dental fluorosis is a condition of the dental hard tissues; it is not a generalized health effect. It is defined as a hypomineralization of enam- el, characterized by greater surface and subsurface porosity than is found in normal enamel, and results from excess fluoride reaching the growing tooth during its developmental stages (Fejerskov et al., 1990~. The staining, which is characteristic of more severe forms of fluorosis, actually develops after took eruption, but is seen only when porous enamel has formed before eruption (Fejerskov et al., 1990~. Surface enamel that exhibits dental fluorosis contains higher concentrations of fluoride than does unaffected enamel, and the fluoride content generally increases with the severity of the condition (Richards et al., 1989~. The fluoride content of crystals in the hypomineralized subsurface layer is low, however, when compared with the content of crystals in the more fully mineralized enamel surface layers (Richards et al., 1989; Yanagi- sawa et al., 1989), and these subsurface crystals appear stunted when examined microscopically. The process of enamel maturation consists of an increase in mineral- ization within the developing tooth and a concurrent loss of early-secreted
Der'tal Fluorosis 23 matrix proteins. Excess fluoride available to the enamel during matura- tion disrupts mineralization and results in excessive retention of enamel proteins. This process has been well-illustrated in animal studies. When rats were given various concentrations of fluoride in their drinking water over a 5-week period, no differences were found in the protein content of fluorotic enamel and control enamel during the secretory phase of enamel formation (Den Besten, 1986~. However, in fluorotic enamel during the early-maturation stage, animals receiving high doses of fluo- ride had more enamel proteins retained. At the late-maturation stage, differences were again less apparent, and only the dental enamel from animals with the highest fluoride intake contained more protein (Den Besten, 1986~. This research indicates that the early-maturation stage is the developmental period when enamel is most sensitive to the effects of fluoride. Other research in various animal models (Richards et al., 1986; Richards, 1990) and humans (Evans and Stamm, 1991a) generally sup- ports the idea that the early-maturation stage is the most critical develop- mental period for dental fluorosis, but fluoride at sufficiently high con- centrations might affect enamel at all stages of its formation (Den Besten and Crenshaw, 1987; Suckling et al., 19X8~. In humans, the clinical signs of severe dental fluorosis (ename! pitting and obvious brown stain) follow the breakdown of the better-mineralized surface layers of enamel shortly after eruption, resulting in variable uptake of mineral in the exposed hypomineralized subsurface lesions ~hylstrup, 1983; Fejerskov et al., 1991~. Some physiological conditions that affect amelogenesis in humans can lead to variations in the clinical appearance of dental fluorosis at similar levels of fluoride intake (Angmar-Mansson and Whitford, 1990~. Cal- cium deficiency and generalized malnutrition are examples of such conditions seen in many developing countries. Any condition that de- creases urinary pH, such as disorders in acid-base balance, can reduce the renal clearance of fluoride and increase the likelihood of dental fluorosis (Whitford and Reynolds, 1979; Ekstrand et al., 1982~. Reten- tion of fluoride in body tissues is increased by high altitudes (Manji et al., 1986a), although residing at high altitudes, in the absence of fluoride, has been found to disrupt amelogenesis and produce a condition that can be clinically confused with dental fluorosis (Angmar-Mansson and Whit- ford, 1990~. Dental fluorosis is a dose-response condition: the greater the intake
24 Health E.ffects of Ingested Fluoride during developmental periods, the more severe the fluorosis will be (Dean, 1942; Eklund et al., 1987; Larsen et al., 1987; Gedalia and Shapira, 19X9; Fejerskov et al., 19901. Evidence from animal studies shows that several patterns of fluoride exposure can disturb amelogenesis. Early research indicated that the development of fluorosis in the con- tinuously growing rat incisor was associated with occasional "spikes" in plasma fluoride concentrations, produced by daily injections, that raised the plasma fluoride concentration above a presumed threshold value (Angmar-Mansson et al., 1976; Myers, 1978~. Later research confirmed that finding (Angmar-Minsson and WhitforcI, 1982) but also showed that relatively constant, slightly elevates! concentrations of plasma fluoride produced by constant infusion in rats (approximately 3 micromoles ~mol)/~) also resulted in enamel fluorosis. A subsequent study, which also employed constant infusion of fluoride in rats, extended the closing period from ~ to ~ weeks. With longer exposure, enamel fluorosis was associated with plasma fluoride concentrations of only i.5 ,umol/L (Angmar-Mansson and Whitford, 1984~. Those results were later con- firmed by Nelson et al. (1989), who found that more fluorosis-type lesions were produced in sheep after long-term administration of low doses of fluoride than after short-term administration of high doses. Angmar-Mansson and Whitford (1985) also reported that a single high dose of fluoride (0.75 milligram (mg) or more of fluoride per kilogram (kg) of body weight) caused enamel fluorosis in rats, even though the plasma fluoride concentrations returned to pre-dose levels within 24 hours. It was hypothesized from those results that the pulse loading (single high dose) and subsequent gradual release of fluoride from bone in the vicinity of the developing enamel result in local fluoride concentrations sufficiently high to disturb amelogenesis. That hypothesis was supported by nuclear microprobe analyses, which showed that the enamel ant! dentin fluoride concentrations were elevated, in a dose-response manner, even 70 days after the single fluoride doses, by which time the rats' incisor teeth would have renewed themselves nearly 2 times (Angmar- Mansson et al., 1990~. Dental fluorosis in humans generally is more severe in teeth that mineralize later in life than in those that mineralize earlier (Larsen et al., 1985, 19X7, 1988; Pablum et al., 19871. That finding is usually attrib- uted to greater ingestion of fluoride by older children comparer] with
Dental Fluorosis 25 younger (although, as a function of body weight, there is often little difference in fluoride consumption between older and younger children). Fluorosis is primarily a condition of permanent teeth; the degree of fluorosis reported in primary teeth is generally much less than that found in permanent teeth (Gedalia and Shapira, 1989~. Although extensive fluorosis of primary teeth has been reported in areas of the world with high amounts of fluoride ingestion (Thylstrup, 1978; Olsson, 1979; McInnes et al., 1982; Nair anti Manji, 1982; Larsen et al., 1985; Mann et al., 1990), it has not been identified as a problem in the Uniter] States. The lower degree of fluorosis in primary teeth was once believer! to be clue to the placenta acting as a barrier to the passage of fluoride from maternal to fetal bloocI, but more recent evidence shows that the placenta acts as only a limite(l barrier to its passage (Gedalia and Shapira, 19891. However, fetal blood concentrations usually are lower than maternal levels. Most fluoride in a tooth's outer enamel layer is clepositec! cluring the enamel maturation period before eruption, a developmental phase that lasts only I-2 years in primary teeth but takes 4-5 years in permanent teeth. The shorter maturation period for primary teeth, adcled to lower fetal blood fluoride concentrations cluring their prenatal clevelopment, is probably the main reason why fluorosis in primary teeth is unusual outside areas of high-fluoride ingestion. DIAGNOSTIC ISSUES IN DENTAL FLUOROSIS Clinical diagnosis of fluorotic lesions has been plagued from the earliest studies by the fact that not all mottling of dental enamel is caused by fluoride. During McKay's initial studies in the early years of this century he referred to this condition as "Coloraclo brown stain," a com- ment both on the geographic location of his investigations and on the severity of the condition he found. In his first national publication on the condition (Black and McKay, 1916), however, it was calle(l "mottled enamel." The term mottled enamel has since evolved to cover a range of dental developmental defects, fluorotic and otherwise;- whereas the term fluorosis more correctly applies only to dental defects of fluorotic origin. The use of these terms, however, is unfortunately far from ·^ unlrorm.
26 Health Effects of Ingested Fluoride The diagnostic problems that can arise when measuring the prevalence of dental fluorosis have been well described in the literature (Fejerskov et al., 1988; Cutress and Suckling, 1990~. Malnutrition, metabolic disorders, and the presence of other dietary trace elements can lead to diffuse, symmetrical markings on the enamel that closely mimic the appearance of fluorosis. Most cases of dental fluorosis are probably identified correctly by experienced examiners, but high and low popula- tion prevalences and individual cases have been reported that are incom- patible with the fluoride histories. In such instances, the likelihood of misclassification seems strong. When a fluoride history is taken concur- rently with clinical diagnosis, some argue that the information from the history biases the diagnostic process. Those issues deserve to be addressed seriously. It must be remem- bered, however, that the risk of misclassification has long been recog- nized as an inherent problem in epidemiological study of any condition. Standard procedures have been developed to minimize the chances of misclassification in data collection ~ilienfeld and Lilienfeld, 1980) and to reduce its biasing effect in statistical analysis (Kleinbaum et al., 1982~. During data collection, misclassification can be minimized by cross- checking with patient records to certify that diagnoses have been cIas- sified correctly ~ilienfeld and Lilienfeld, PESO). For dental fluorosis classification, taking a fluoride history seems to be an appropriate part of a difficult diagnosis. Of course, some diagnostic errors will still be made, but probably no more than are made in clinical examinations for caries, loss of periodontal attachment, soft-tissue lesions, or a host of meclical conditions. The alternative approach is to avoid fluorosis misclassification in the data collection by recording less well-def~ned entities, such as "enamel opacities" or "developmental defects." Whether or not that approach reduces misclassification, it can yield data that are of limited use in answering research questions. For example, it is not easy to interpret the finding that "some type of defective enamel" was found in 50.~% of a group of schoolchildren, and "white diffuse or patchy" opacities were found in 10.2% of them (summer et al., 1990~. The latter description best fits that of dental fluorosis, but the authors of that report do not use the word fluorosis at all. In another study, "at least one tooth with defective enamel" was found in 63% of children seen, and 4.4% of them showed "diffuse patchy opacities" (Suckling and Pearce, 19841. The latter description is closest to that of dental fluorosis of all the descrip
Dental Fluorosis 27 tions given in that report, but again the reader is left to make that inter- pretation. Others compare the proportion of "blemishes" found in people in fluoridated and nonfluoridated areas (Dooland and Wylie, 1989) and the trends in prevalence of "mottling" since the introduction of fluoride toothpastes (weeks, 1990~. Even though the problem of misclassifying dental fluorosis is recog- n~zed during clinical assessments, determination of risk factors for fluorosis seems more likely to be successful when attempts are made to measure the condition directly. Russell's guidelines for distinguishing between enamel opacities of fluoride and nonfluoride origin, now more than 30 years old (Russell, 1961), are still a valid diagnostic guide. Taking a fluoride history, when practical, is also a standard and appropri- ate epidemiological procedure that should not bias the data collected. INDEXES FOR DENTAL FLUOROSIS This section provides a brief description and critique of the most common indexes used to grade dental fluorosis. The detailed clinical criteria and scoring systems for the three most frequently used indexes (Dean's, Thylstrup and Fejerskov's, and the Tooth Surface Index of Fluorosis) are given in Appendix I. The first index to grade dental fluorosis was developed by Dean in 1934; he called it the Fluorosis Index, but it is usually referred to as Dean's index. Dean had been assigned by the U.S. Public Health Service to investigate what was then the newly identified condition of dental fluorosis and to determine if it was a public-health problem requiring action. His first index (Dean, 1934) was fairly arbitrary and consisted of a seven-point ordinal scale from normal to severe. After some years of use, Dean modified the index to form the six-point ordinal scale that is still used today (Dean, 1942~. The scores from Dean's index are based on the two worst-affected teeth in the mouth and are derived from the whole tooth rather than the worst-affected surface. Water research by Thylstrup and Fejerskov (197X) states that all surfaces of an affected tooth should be affected equally.) In the examination, teeth are not dried. Dean's index has stood the test of time, but even though it is adequate for broad definition of prevalence and trends, it is not sufficiently sensitive at both ends of the scale for analytical research. Dean's index inevitably was modified to meet different conditions in
28 Health Effects of Ingested Fluoride different parts of the world. The ThyIstrup-Fejerskov (TF) index is an extensive modification of Dean's with a strong biological basis (Thylstrup and Fejerskov, 1978~. The lO-point ordinal TF index is more sensitive than Dean's at the high end of the severity scale, and it appears to be more sensitive at He mild end of the scale because its application calls for teeth to be dried, which makes the mildest fluorosis more likely to be detected (Granath et al., 1985; Cieaton-Iones and Hargreaves, 19901. Indeed, the lowest scores in the TF index reflect such mild fluorosis that some have questioned whether those categories should be inclucled in the index as fluorosis categories (O'Mullane and CIarkson, 1990~. The TF index has been shown to be a valid indication of fluoride content of fluorotic enamel, although teeth scored in the first three categories of the index hac! fluoride concentrations that were similar to those in normal enamel (Richards et al., 1989~. The TF index might be too sensitive in those categories, which slider only slightly in their clinical appearance, for purposes other than analytical epidemiology (CIarkson, 19891. The Tooth Surface Index of Fluorosis (TSTF) was developed by re- searchers at the National Institute of Dental Research in the early 198Os (Horowitz et al., 19841. It is intended to be more sensitive than Dean's index by ascribing a score to each unrestored surface of each tooth (rather than a single tooth score to the two worst-affected teeth in the mouth), by eliminating Dean's category of "questionable," and by provi(i- ing greater range at the high end of the severity scale. The tooth is not ctried ciuring TSIF examination. Whether it achieves greater sensitivity with these means has not been conf~rmecI, although it probably provides a more valic] picture of whole-mouth severity than does Dean's inclex. Its use of staining as a criterion has been criticized (Fejerskov et al., 1990), because staining is a post-eruptive phenomenon that is a function of a person's dietary habits as well as degree of enamel porosity. The Fluorosis Risk Index (FRI) was developed to relate age-specific fluoride exposure to development of dental fluorosis (PencIrys, 19901. It divides the surfaces of permanent teeth into two developmentally related groups of surface zones, which begin formation either during the first year of life or cluring the third to sixth years. It was developed specifi- cally for use in case-control studies ant} thus has clear rules for categoriz- ing subjects as cases, controls, or neither, depending on the distribution of fluorosis found on designated zones of tooth surfaces. The ability of FR} in a case-control study to relate fluorosis to enamel developmental
Dental Fluorosis 29 periods permitted identification of risk factors that other indexes could not do (Pendrys and Katz, 1989~. Its use is best restricted to analytical epidemiology. The Developmental Defects of Enamel Index (DDE) was developed by a working group of the Federation Dentaire Internationale, and, as its name implies, it was intended to be more than a dental fluorosis index (FDI, 1982~. One of the reasons given for developing the DDE was that "Classifications based on etiological considerations are premature because only a few defects can be assigned an etiology" (FDT, 1982~. The diagnostic problems described in the previous section affect the DDE because it assigns codes for all types of enamel opacities and thus has made it difficult for researchers to classify fluorosis as distinct from other enamel defects. Data analysis in the DDE is also complicated (CIarkson, 1989~. A modified version, intended to make identification of fluorosis simpler, has been used in a national survey of Ireland (CIarkson and O'MulIane, 1989~. In this survey, diffuse opacities were found to be the discriminating factor between fluoridated and nonfluoridated areas. All the indexes have strengths and deficiencies, and choice of index should be determined by the purposes of the study. They are not the only systems for classifying dental fluorosis that have been proposed (Al- Alousi et al., 1975; Butler et al., 1985a), but they are the ones most in use at present. The major problem that arises from use of multiple indexes is that direct comparison of prevalence and severity is difficult when classifications of cases of fluorosis vary according to the index used. The review carried out by the U.S. Public Health Service (PHS, 1991) even considered reports in the literature according to the index used, a conservative approach but one that is hard to criticize purely on scientific grounds. In this review, we attempt to pool results but focus comparisons on prevalence and broad categories of "mild to very mild" and "moderate to severe." DENTAL F.LUOROSIS AND FLUORIDE INTAXE Dean's initial research established I.0 mg/L as the approximate con- centration of fluoride in drinking water that best preventer] caries while keeping unsightly dental fluorosis to a minimum (Dean, 1942), but in
30 Health EfFects ofIngested Fluoride reaching that conclusion, Dean did not include the likelihood that more water is consumed by people in warm climates than in coo! climates. After Dean's work, the "optimal" concentration of fluoride in drinking water for the United States was established by the U.S. Public Health Service to be between 0.7 and I.2 mg/L, depending on mean temperature of the locality (PHS, 1962~. Regions of the country with warm climates (e.g., Arizona and Texas) use 0.7 mg/L as their optimal concentration, and coo} regions (e.g., Minnesota and New England) use I.0-~.2 mew. The recommended water fluoride concentrations for ranges of mean temperature in the United States are shown in Table 2-~. The range of 0.7-~.2 mg/L was established after a series of elegant studies in the 1950s that related fluoride concentrations in drinking water to fluorosis prevalence and severity in different climatic regions in the United States (Galagan, 1953; Galagan and Lamson, 1953; Galagan and Vermillion, 1957; Galagan et al., 1957~. The results of those studies were confirmed by further research a decade later (Richards et al., 1967~. As was the case wig all fluoride research at Hat time, drinking water was virtually He only source of measurable fluoride. (Foods usually contained only trace amounts of fluoride.) Although fluoride is no longer considered an essential factor for human growth and development (see NRC, 1989), many believe that there is an optimal dose of systemic fluoride for maximal benefit against caries. (This is not the same subject as the "optimal concentration" of fluoride TABLE 2-1 Water Fluoride Concentrations Recommended for Vanous Climatic Conditions Annual Average Maximum Daily Air Temperatures, °F (°C) 50.0-53.7 (10.0-12.0) Recommended Control Limits, mg/L Lower Optimum Upper 0.9 1.2 1.7 53.8-58.3 (12.1-14.6) 0.8 1.1 1.5 58.4-63.8 (14.7-17.7) 0.8 1.0 1.3 63.9-70.6 (17.8-21.4) 0.7 0.9 1.2 70.7-79.2 (21.5-26.2) 0.7 0.8 1.0 79.3-90.5 (26.3-32.5) 0.6 0.7 0.8 Source: PHS, 1962.
Dental Fluorosis 31 in drinking water and should not be confused with it.) But because the controversy regarding pre-eruptive and post-eruptive fluoride action still exists, this issue deserves some scrutiny. The concept of an optimal dose goes back to the early days of fluoride research in dentistry. In 1943, the normal daily fluoride intake of chil- dren I-12 years old was estimated to be 0.4-~.7 ma, which provided an average intake of fluoride at 0.05 mg/kg of body weight per clay (McClure, 1943~. Actual fluoride intake for an individual depended on age, diet, and fluoride content of water. That estimate somehow evolved into a recommendation (Farkas and Parkas, 1974) and then to apparent acceptance of 0.05-0.07 mg/kg per day as an optimal dose (Ophaug et al., 1980a). Despite its dubious genesis, that dose might be a fair estimate, based on empirical evidence, of the upper limit for fluoride intake in children to minimize fluorosis (Burt, 1992~. If all fluoride intake comes from drinking water, that dose for a child weighing 10 kg (an average I-year- old) would be ingested in 0.5-0.7 ~ of water fluoridated at I.0 mg/~. For a child weighing 22 kg (an average 6-year-old), it would be ingested in i. I-~.S ~ of water. Because the scientific base is weak, however, the range of 0.05-0.07 mg/kg should not be referred to as an optimal dose, and it should not be considered more than a guide to the upper limit of intake for minimizing fluorosis. The intake of fluoride that leads to clinically detectable dental fluoro- sis, relative to body weight at different stages of growth, still requires considerable clarification. Forsman (1977) stated that a daily intake of 0. ~ mg/kg was sufficient to cause dental fluorosis, an estimate that was later revised downward to 0.04 mg/kg (Barium et al., 1987~. Studies in Kenya, where naturally occurring high-fluoride drinking water is common in the Rift Valley, led to an even lower estimate of 0.03 mg/kg (phylum et al., 1987~. Other sources of fluoride were not accounted for in those observations, and none of the estimates specified the period of childhood in which the intakes were most critical. The estimates of 0.03-0.04 mg/kg were for those in the lowest categories of the TF index, in which, as noted earlier, the fluoride content of fluorotic lesions is similar to that of normal enamel. Further research is necessary to clarify the relation between fluoride intake in childhood and development of dental fluorosis. Recent estimates of daily intake of fluoride from food and drink by North American children up to 2 years of age are 0.01-0.16 mg/kg in areas
32 Health Effects of Ingested Fluoride without fluoridation and 0.03-0.13 mg/kg in areas with fluoridation (Burt, 1992). pRE~rA~ENcE OF DENTAL F~UOROSIS Because dental fluorosis is a dose-response condition (Myers, 1983), severity ranges from barely discernible, even to a trained observer, to the most severe manifestations of stained and pitter! enamel. Figure 2-1 demonstrates this range, beginning with unaffected enamel in Figure 2-lA and encling with severe dental fluorosis in Figure 2-IF. The discussion on the prevalence of dental fluorosis, which follows, refers to these grades of severity. . ~ 1~ ·- _I ~ ~ :~ ~ ~' Severe dental fluorosis Figure z-le and z-1~) ~s e'~ut;~c ~ Ala U1 the world where extremely high concentrations of fluoride occur naturally in drinking water. In subtropical zones of India, for example, fluoride concentrations as high as 15-16 mg/L have been recorded (Iolly et al., 1968; Chandra et al., 1980; Subbaredcly and Tewari, 19851. In addition to severe dental fluorosis, skeletal fluorosis is not uncommon in such areas (jolly et al., 1968~. East Africa and parts of the Middle East also have excessively high concentrations of fluoride in drinking water. In Kenya, fluoride concentrations of over 40 mg/L have been recorded in drinking water; in one study, over 20% of 1,290 boreholes in the Great Rift contained fluoride at more than 5 mg/L (Manji and Kapila, 1986a). As would be expected, severe dental fluorosis is endemic in many parts of those regions (Mailer et al., 1970; Olsson, 1979; Walvekar and Qureshi, 19X2; Wenze! et al., 1982; Manji et al., 1986b; Manji and Kapila, 1986b; Chibole, 1987; Haimanot et al., 1987; Mann et al., 1987~. FIGURE 2-1 (Appears on the next two pages) (A) No fluorosis detectable. (B) Very mild fluorosis, most noticeable at the tips of the upper and lower ~L. -~U ^~ 1_~ /~ MilA Hlinro~i.c Diffuse white striations visible over much of the upper incisor teeth; TSIF score 3. (D) Moderate fluorosis. Some brown stain visible among the white striations on the upper central incisors; TSIF score 4. (E) Severe fluorosis. Discrete pitting of the enamel of the upper and lower front teeth; TSIF score 5. More severe confluent pitting can be seen on the lower bicuspid tooth (the lower tooth at the end of the arch); TSIF score 7. (F) Severe fluorosis. Discrete pitting and stain visible; TSIF score 6. 1nt;lNOF 1~111. 1~11' ~ ~ -I. ~& ~ a~
Dental Fluorosis 35 Dean's research during the 1930s and 1940s, conducted in areas with varying amounts of naturally occurring fluoride in drinking water, aimed to establish fluoride concentrations that represented the best balance between low occurrence of caries and acceptable level of dental fluorosis. From his extensive field research, Dean concluded that fluoride at I.0 mg/L of drinking water was the "minimal threshold of endemic dental fluorosis" and noted that, at I.0 mg/L, 10-12% of permanent-resident children showed the mildest forms of fluorosis (Figure 2-IB and 2-!C), mostly in the bicuspids and molars (Dean et al., 1941~. Dean's research resulted in acceptance of fluoride at I.0 mg/L in drinking water as the most appropriate concentration for most of North America, until the climate-related range of 0.7-~.2 mg/L was adopted by the U.S. Public Health Service in 1962. The early supplemental fluoridation studies (fluoride at I.0 mg/L~) in Grand Rapids, Michigan, and Newburgh, New York, conducted at a time when drinking water was virtually the only important source of fluoride, shower! that dental fluorosis generally was as prevalent and severe as it was in communities with naturally occurring fluoride at the same concentration (Ast et al., 1956; Russell, 1962~. To establish a "standard" against which the present-cay prevalence of dental fluorosis can be assessed, Figure 2-2 shows the distribution of fluorosis in 10 of the many communities studied by Dean in the 1930s and 1940s. The 10 were chosen to demonstrate effects from a range of water fluoride concentrations, and all had more than 50 children of ages 9-14 years in the groups studied. All data were collected by Dean and categorized with Dean's index. Prevalence of fluorosis can be seen to be directly related to fluoride concentration of drinking water up to ap- proximately 4.0 mg/L, after which it tended to level out. Moderate-to- severe dental fluorosis began to be seen at about I.9 mg/L, and its prevalence rose with increasing fluoride concentrations. It is worth noting that even where drinking water contained fluoride at only 0.4 mg/L, there was some fluorosis recorded, and the relations, although direct, were by no means linear (Dean, 19421. The research of the 1930s set the stage for controlled addition of fluoride to drinking water at about I.0 mg/L and was followed by studies through the 1940s and 19SOs devotee! to evaluating the initial controlled fluoridation projects. After the mid-1960s, however, research into dental fluorosis in the United States dropped off sharply. It was revived only in the 19SOs, after Leverett cautioned that fluoride exposure from wide- spread use of an increasing number of sources of fluoride should be
36 Health Effects of Ingested Fluoride 100 80 a) `' 60 40 20 lo ~ Moderate/Severe · Mild/\lery Mild . 0.4 0.9 1.3 1.9 2.2 2.6 3.9 4 4.4 7.6 Water Fluoride Concentrations FIGURE 2-2 Prevalence (percent) and seventy of dental fluorosis in 10 com- munities in the United States in the 1930s and 1940s by fluoride concentration (mg/L) in drinking water. Source: Dean, 1942. monitored Everett, 1982~. The new interest in monitoring the impact of fluoride through the 198Os and 199Os yielded evidence that the preval- ence of dental fluorosis had increased since Dean's time in both fluori- dated and nonfluoridated areas. In the National Survey of Dental Caries in U.S. School Children in 19X6-1987, overall prevalence of dental fluorosis was reported to be 22.3% in children ages 5-17 years (Brunette, 19891. Nearly all the eases were mild to very mild. Perhaps the most noteworthy aspect of this survey (to date reported only in abstract form) was that prevalence was IS.5% in 17-year-olds and 25.~% in 9-year-olds, a cohort comparison that suggests that childrens' exposure to fluoride increased during the 1970s. (The differing prevalence by age has been suggested to be at least partly due to decreasing fluorosis severity in the same tooth over time, a change attributed to post-eruptive mineralization and to abrasion of affected incisal edges. The issue of whether the milder forms of fluorosis do in fact diminish over time requires further research.) Although dental fluorosis is still more prevalent in fluoridated than in nonfluoridated communities, its prevalence has increased proportionately more in the
Dental Fluorosis 37 nonfluoridated areas since Dean's work Everett, 1986; Bagramian et al., 1989; Kumar et al., 1989; Williams and Zwemer, 19901. Several detailed reviews of the literature (Szpunar and Burt, 1987; Pendrys and Stamm, 1990) comparing fluorosis data over time, in addi- tion to other recent research, concluded that the prevalence of dental fluorosis reported in optimally fluoridated areas (both natural and added) in recent years ranged from 8% to 51%, compared with 3% to 26% in nonfluoridated areas. Those ranges consist of all degrees of severity, although 90% or more of fluorosis cases recorded in the United States are in the mild-to-very-mild category. More recently, a prevalence of 80.9% was reported in children 12-14 years old in Augusta, Georgia, the highest prevalence yet reported in an optimally fluoridated community in the United States (Williams and Zwemer, 1990~. Again, most of the fluorosis was mild to very mild, but moderate-to-severe fluorosis was found in 14% of the children. Several factors might be contributing to this high prevalence. For example, Augusta is described as optimally fluoridated, but the water fluoride concentrations of 0.9-~.2 mgIL seem excessively high for that climatic region, and over 80% of participants were reported to have used fluoride supplements in childhood (clearly inappropriate in a fluoridated area). Because only 33.4% of eligible children in that study responded to the researchers' initial contact, there is also the possibility of selection bias in the examined group. Pendrys and Stamm (1990) estimated that in the 1930s, residence in an optimally fluoridated area carried about an I8-fold increase in the risk of dental fluorosis, whereas the current increase in risk is about 2-fold. That reduction is largely attributable to an increase in fluorosis in non- fluoridated areas, where fluoride supplements, beverages processed with fluoridated water, and inadvertent swallowing of fluoride toothpaste represent sources of fluoride that were not present in the 1930s. As can be inferred from the ranges of prevalence already given, the extent of dental fluorosis in a community cannot be estimated from water fluoride concentrations alone. To illustrate, Figure 2-3 shows the preval- ence of dental fluorosis in the early 1980s and its distribution, using Dean's index, of categories between mild to very mild and moderate to severe in 16 communities in Texas with varying water fluoride concentra- tions (reported only as a function of optimal concentration, presumably 0.7 mug/. The number of examiners in the study was not reported. Although direct comparisons wig Dean's data must be made cautiously
38 Health Effects of Ingested Fluoride 100 80 60 a) 3; 4o ~ _ ~ - _ <:~ cs3 cub ~ Moderate/Severe ~ Mild/\/ery Mild ,~ ,~ ,~ ~3 ,~3 ,~ At ,, ~A ,~9 ,,~ ~,~ Water Fluoride Concentration (x Optimum) FIGURE 2-3 Prevalence (percent) and severity of dental fluorosis in 16 com- munities in Texas in the 1980s by fluoride concentration in drinking water. Source: Segreto et al., 1984. (and there are some curious inconsistencies in the Texas data), prevalence does appear higher in the 198Os than in the 1930s. Prevalence can be seen to vary from X.7% when water contained 0.3 times the optimal concentration to 94.7% when water contained 4.3 times the optimal concentration (Segreto et al., 19X4~. The inconsistencies between dental fluorosis prevalence and water fluoride concentrations, seen in Figure 2- 3, are difficult to explain. Even though examiner variability is likely to be a factor, the inconsistencies cannot be attributed solely to those varia- tions because the inconsistencies are seen in most series of that kind (e.g., Dean's data in Figure 2-2) and in the ranges of prevalence given in the previous paragraph. They are most likely a result of different individual fluoride intakes, not all of which can be detected without a significant increase in research effort, and variation in individual biologi- cal responses to similar fluoride intakes. Results of semi-Iongitudinal assessments of dental fluorosis in seven communities in Illinois, conducted by researchers from the National Institute of Dental Research (NIDR), showed the prevalence of fluorosis
Dental Fluorosis 39 and its changing distribution in recent years (Heifetz et al., 1988~. Those communities were chosen for the varying concentrations of naturally occurring fluoride in their drinking water. Figure 2-4 shows the preval- ence and severity of dental fluorosis in children who were 13-15 years old ant] who were permanent residents in Dose communities in 1985. Two examiners used the TSIF index, in which graces I-3 generally correspond to the mildest forms of dental fluorosis and grades 4-7 reflect moderate-to-severe forms. Repeat examinations demonstrated reasonable agreement between the examiners, so their results were pooled. Overall, dental fluorosis prevalence was noticeably higher at the lower fluoride concentrations than Dean had recorded but differed little at 3-4 times the optimal concentration. Moclerate-to-severe fluorosis might be even less prevalent, according to these data, than Dean recorded about 50 years earlier. In their first report, the NTDR researchers concluded that the preval- ence of clental fluorosis in children 8-16 years old h act not increased sig- nificantly since Dean's surveys in the 1930s (Driscoll et al., 1986~. That | O Moderate/Severe ~ Mild/Very Mild | 100 80 a, ~ 60 a) a) 40 20 o Optimum 2 x 3 x 4 x Water Fluoride Concentration (x Optimum) FIGURE 2-4 Prevalence (percent) and severity of dental fluorosis in Illinois in 1985 by fluoride concentration in drinking water. Source: Heifetz et al., 1988.
40 Health Effects of Ingested Fluoride conclusion was based on data collected in 1980. However, significant increases in prevalence between 1980 and 1985 were reported in cohorts 13-15 years of] in the Illinois communities (Heifetz et al., 198X), al- though they could not be discerned in cohorts 8-10 years old over the same period. Relating age to dental fluorosis and tooth calcification, the authors concluded that fluoride intake (from all sources) had increased from 1970 to 1977 but had not increased much since then. To focus on Me extent of dental fluorosis at higher water fluoride concentrations uncler modern conditions, Figure 2-5 displays data from three studies in communities where water fluoride concentrations were twice the optimal concentration or higher. Overall, prevalence is high, and most variation seems to come from the proportion of dental fluorosis graded moderate to severe or mild to very mild. It is difficult to specify how much of the variation is due to differences among examiners, vary- ing intakes of fluoride in water and in other sources, differences in other aspects of diet or physiology, individual biological variations, or some combination of those possibilities. loo 80 a' '' 60 a) 40 20 o ~ ~ Moderate/Severe · Mild/\/ery Mild ~ 2 2.3 2.5 2.7 2.7 2.7 2.9 3 3.1 Water Fluoride Concentration (x Optimum) 4.3 5 FIGURE 2-5 Prevalence (percent) and severity of dental fluorosis since 1980 for selected communities in the United States with above-recommended fluoride concentrations in drinking water. Sources: Segreto et al., 1984; Eklund et al., 1987; Heifetz et al., 1988.
Dental Fluorosis 41 A 199 ~ report from PHS compiled the results of independent investiga- tions conducted during the 19SOs on dental fluorosis in 24 cities and compared them with a series of PHS surveys conducted during the late 1930s and early 1940s in 21 cities (Table 2-2~. That comparison showed that the prevalence of dental fluorosis, most of it mild to very mild, had increased, although the modern~ay effects of fluoride from sources other than water were not controlled for. The 19SOs data showed that the mean prevalence of dental fluorosis in four cities with optimally fluori- dated water supplies was about 22% (17% very mild, 4% mild, 0.~% moderate, and 0.~% severe). In another city with a water fluoride concentration in the range of .-2.2 mg/L, dental fluorosis prevalence was 53% (23% very mild, 17% mild, 8% moderate, and 5% severe). In two other cities with water fluoride concentrations greater than 3.7 mg/L, prevalence was about 84% (25% very mild, 27% mild, 19% moderate, and 14% severe). The data in the PHS report also showed that the greatest relative increase in fluorosis prevalence since the early studies was in communities with very low water fluoride concentrations, demonstrating the influence of sources of fluoride other than water. Those sources make it difficult to estimate fluoride exposure; they repre- sent a confounding factor in studies of the relation between fluoride exposure and dental fluorosis. RISK FACTORS IN DENTAL FLUOROSIS Dental fluorosis is a function of total fluoride intake during critical dental developmental periods, and in modern conditions, fluoride is ingested from numerous sources in addition to drinking water. Research efforts to measure the quantities ingested from all such sources are often frustratingly imprecise, because quantifying fluoride intake from current and past use of water, food, and toothpaste, together with past intake from supplements or infant formula, can be extremely difficult. No tissues of the body can be measured for lifetime intake of fluoride. Measurements in plasma might be the best but can be affected by changes in recent intake. As a relatively invasive procedure, it is also not easy to use in field studies. Bone obviously cannot be biopsied from volunteer study participants. Nail clippings reflect only fairly recent fluoride intake and can be readily contaminated. Despite these measurement problems, risk factors for dental fluorosis
42 Health Effects of Ingested Fluoride TABLE 2-2 Percent Prevalence of Dental Fluorosis by Clinical Classification and Concentration of Water Fluoride from Dean's 1940 21-City Survey and 1980's Sunrey Using Dean's Index Total No. Cities and Studies Water Fluoride, 1940s 1980s <0.4 Prevalence, Pa Mean Sample Size per City and Study Very Mild 1940s 1980s 1940s 1980s 10 5 360 326 0.9 0.7 4.4 3.0 0.4~.6 3 1 427 126 5.0 ~2.4 1.4 0.7-1.2 4 4 270 471 12.3 + 17.7 + 2.3 16 1.3-1.7 1 3 477 175 27.0 ~19.5 + 4.2 4.1 1.8-2.2 2 1 222 143 35.1 ~23.1 7 2.3-2.7 1 5 4~)4 174 42.1 40.5 7 2.8-3.2 0 3 124 26.2 16 3.3-3.7 0 0 >3.7 0 2 163 24.8 11 Total 21 24 aMean it standard deviation. Note: The 21 cities represented in the 1940s column are those cities sur- veyed by Dean in the 1940s. The 24 cities represented in the 1980s column are those cities surveyed by different investigators using Dean's index in the 1980s. The means and standard deviations are derived from the cities classified by respective water fluoride concentrations. Source: PHS, 1991. have been identified from epidemiological study, but it is still difficult to rank their importance with any certainty. One such risk factor is ob- viously a high fluoride concentration in drinking water (Szpunar and Burt, 19881; even minor adjustments in water fluoride concentrations can lead to significant changes in the prevalence of clinically detectable fluorosis (Evans and Stamm, 199Ib). Other risk factors are ingestion of
Dental Fluorosis 43 Prevalence, %. (Continued) Mild Moderate Severe Total 1940s 1980s 1940s 1980s 1940s 1980s 1940s 1980s 0 2.2 + 0 0.1 + 0 0 0.9 ~6.4 + 2.8 0.3 0.7 2.6 0.6 + 0 0 0 0 0 5.6 + 2.4 0 3 1.2 1.4+ 4.4+ 0 0.8+ 0 0.1 + 13.6+ 22.2+ 0.4 2.0 0.3 0.3 2.0 14 3.1 + 5.6 ~0 0.7 + 0 0 30.2 + 25.7 + 1.1 4.7 0.6 4.3 9 7.5 + 2 16.8 1.1 + 8.4 0 4.9 44.0 + 53.2 0.1 6 21.3 29.5 + 8.9 8.4 + 5 1.5 0 73.8 78.5 + 5 9 30.0 + 15.0 + 2.8 + 5 74.0 + 11 16 22 27.8 + 19.3 + 13.9 + 83.4 + 4 17 12 16 fluoride (both intentional and inadvertent from sources other than drink- ing water. Indeed, most dental researchers (Horowitz, 1991; Rozier, 1991; Szpunar and Burt, 1992) believe that the best approach to stabiliz- ing the prevalence and severity of dental fluorosis is to control fluoride ingestion from foods, processed beverages, and dental products rather than reduce the recommended concentrations of fluoride in drinking water. A large number of studies have concluded that fluoride supplements are a risk factor for dental fluorosis (Holm and Andersson, 1982; Suck ling and Pearce, 1984; Hellwig and Klimek, 1985; Bohaty et al., 1989; Dooland and Wylie, 1989; Kumar et al., 1989; Larsen et al., 1989;
44 Health Effects of Ingested Fluoride Pendrys and Katz, 1989; Woolfolk et al., 1989; Woltgens et al., 1989; Holt et al., 1990; Small et al., 1990; Riordan and Banks, 1991; Lalumandier, 19921; however, the relation shown was weak in some of the reports. Other studies have failed to show such a relation (Butler et al., 1985b; Bagramian et al., 1989; Stephen et al., 1991~. Fluoride supplements are used widely (Brunelle and Carlos, 1990) and often are prescribed inappropriately (Pendrys and Morse, 1990; Levy and Muchow, 1992~. Some of the compliance data are imprecise, however, because use is usually documented retrospectively. It is not clear wheth- er it is solely the misuse of supplements that results in dental fluorosis or whether use at recommended dosages produces the condition (Workshop Report, 1992~. In either case, the appropriateness of the recommended supplementation schedules should be considered. The swallowing reflex is not fully developed in children of preschool age, and their inadvertent swallowing of fluoride toothpaste has been identified as a risk factor (Hellwig and Klimek, 1985: Osuii et al., 19X8; Pendrys and Katz, 1989; Milsom and Mitropoulos, 1990; Lalumandier, 19921. However, the fluorosis reported was often very mild and barely discernible. In addition to toothpaste, prolonged use of infant formula in the fluoridated area of Toronto, Ontario, was identified as a risk factor for dental fluorosis (Osuji et al., 19881. High socioeconomic status also emerged as a strong risk factor in a well-conducted case-contro' study (Pendrys and Katz, 1989), but that finding has not been confirmed in other studies (Bagramian et al., 1989; Hamdan and Rock, 1991~. Fluo- ride in foods and beverages processed with fluoridated water has long been suspected as a risk factor but has not been clearly demonstrated. Unexpectedly high fluoride concentrations in particular foods and bever- ages (Clovis and Hargreaves, 1988; Burt, 1992; Pang et al., 1992), however, might stimulate further research in this area. Although the subject has received little attention, some data suggest that dental fluorosis is more prevalent among African-Americans than among other races or ethnic groups in the same community. Russell (1962), in the Grand Rapids fluoridation study, noted that fluorosis was twice as prevalent among African-American children than white children. In He Texas surveys in the 1980s, the odds ratio for African-American children having dental fluorosis, compared with Hispanic and non-His- panic white children, was 2.3 (Butler et al., 1985b). Dental fluorosis also tended to be more severe among African-American children than white children in the Georgia study (Williams and Zwemer, 1990),
Dental Fluorosis 45 although the difference was not statistically significant. In Kenya, prevalence and number of severe cases were unexpectedly high when related to fluoride concentrations in drinking water (Manji et al., 1986c), although nutritional factors could have confounded these results. The reasons for these findings are unknown and do not appear to have been explored further. THE RELATION BETWEEN DENTAL F1UOROSIS AND CARIES Dean's studies of this relation in the 1930s showed a sharp reduction in caries prevalence when communities were ranked from the lowest water fluoride concentrations (virtually zero) to approximately ~ .0 mg/~. His data also indicated that caries prevalence leveled out when communi- ties with water fluoride concentrations above ~ .0 mg/L were rank-orderec] (Dean, 1942~. On the other hand, caries prevalence was observed to increase when fluoride concentrations were such that severe dental fluoro- sis was common and the enamel of affected individuals was friable and liable to fracture (Grobler et al., 19X6~. Other data on the relation are inconsistent. Some studies have found that data follow the I-shaped path shown in Figure 2~: with increasing fluoride concentrations, caries prevalence diminishes to a certain point and then increases again. How- ever, a different relation is seen in Figure 2-7, in which caries experience among adults in Lordsburg, New Mexico, which had 5 times the optimal fluoride concentration in drinking water, was below that found in the neighboring community of Deming, which had an optimal concentration. Differences were found in the amount of dental treatment received be- tween the two communities, however, a factor that could have influencer! the results. An earlier Texas study also reported that caries prevalence among children 12-15 years old continued to diminish even when com- munity fluoride concentrations were 6-8 times the optimal concentration (Englander and DePaola, 1979~. These contradictory findings are dif- ficult to explain and merit further research. CONCLUSIONS The data show that the prevalence of dental fluorosis, nearly all of it
46 Health Effects of Ingested Fluoride 6 5 In IL 3 a) ~ 2 ~1980 - 19851 o Optimum 2x optimum 3x optimum Water Fluoride Concentration 4x optimum FIGURE 2-6 Canes experience of children 13-15 years of age in Illinois in 1980 and 1985 relative to the fluoride concentration in community drinking water. DMFS = decayed, missing, and filled surfaces. Source: Heifetzet all, 1988. mild to very mild (Figure 2-IB and 2-IC), rises with increasing fluoride concentrations in drinking water. The data also show fairly consistently that a small, though measurable, proportion of a population exhibits moderate-to-severe dental fluorosis with .-2.0 times the optimal con- centration of fluoride in drinking water, and this proportion generally increases with increasing concentrations of fluoride. However, the data are not consistent enough to permit a firm definition of the relation between moderate-to-severe dental fluorosis and water fluoride concentra- tions. In addition, other uses of fluoride, independent of water fluoride concentrations, clearly affect the prevalence of dental fluorosis. Develop- ment of a firm public-policy recommendation is also inhibiter! by lack of knowledge of the public's perception of less-than-severe fluorosis (Figure 2-1B, 2-IC, and 2-lD). Public policy on use of fluoride to promote oral health should be aimed at keeping dental fluorosis prevalence as low as possible relative to the benefits of caries control, a classic public-health tracle-off. When
Dental Fluorosis 47 12 10 Cl) 8 2 a) - 2 4 2 | HI Lordsburg ~ Deming | l Rl 1~1 R ~ I_ RAIR R SIR My ~ ........ ,,,., , ,,,,, , . , , ,.,,.,.,,,.,,.,,,, . , , ,, , ,,,,,~ ~ 1 , ,.,,,., ,, ,,,,,,,, , , ,.' , R IlR I , ~ Rtt. l ! ~ , R 89 l i t l R ~ l l -------- ---------- -------- ---1 1 .,,, ,,., ~ ~ _ 1 2 ' ~. 1 1 . : -i 1 ~ ~ 1~ at l 1 .......................... 1 6 ~ · 1 ~ l 1 1:. . 2::::2 2 """"'-. ": _ 1:.-.2.~ 2 : · 2 2 2:2 - :s , 1 ~ ::2 ::::::::- ::.::::2'-.- 1 1:. :::::: ~ : F - - .: -.: :s , I :::::::::::::::.::::--- :: -.- I 1: ".2 2::"2 "::-""2::''-: _ 1:::-"2. --: -:::: - 2-:: - :s , 1 ::: ::::::--::::-.2.2.- :::::::: 1 --- 122222'"2'''"'' I'''' --- 1'''''"''''''"' ~1 .. ,, ,, , ,,,,,,,, , '''''''''''''''''''1 1:::-2:::::: :::::: :::: :.::: _ 1:.2-2." :::: :2 2 - 2-: - 2:.: :s , 1 :::::::2- -::- ::2 - -.2 - 2-:::: 1 2c,,'2:,2f',~,2-~->,-5, ·., ,, ,1 1 '22--- ~1"''''""'''! 1 ............. .1 1:::::::2 2 ::::2::::.'.'.:::::::: ~1:'::: :: .::: :.: .- . :e 1 , ::: :,.: :.-:::::::: 1 1''-''"' ~1'''''''''"! 1 . 1 o 27-40 41-50 51-65 Age Groups ,,,,,,,,,,,,,....,....~ FIGURE 2-7 Caries experience of adults 27~5 years of age in Deming (flu- onde at 0.7 mg/L) and Lordsburg (fluoride at 3.5 mg/L), New Mexico, in 1984. Source: Eklund et al., 1987. drinking water is the only source of fluoride, the evidence supports the conclusion that water fluoridation at currently recommended concentra- tions results in prevalence of mild-to-very-mild dental fluorosis of about 10% and very little severe fluorosis. At twice the recommended con- centrations, the prevalence of moclerate-to-severe clental fluorosis is small but measurable. At higher concentrations in drinking water, the preval- ence increases, although limited evidence shows that the extent ant! distribution of dental fluorosis at 4 times the optimal concentration is not much higher than that at 2 times the optimum. At 5 times or more, however, the prevalence of moderate-to-severe dental fluorosis is substan- tially higher. Interpretation of current data is difficult because of exposure to fluo- ride from sources other than drinking water. In the modern U.S. envir- onment, people are exposer] to fluoride from food, beverages, toothpaste, and a variety of prescribed or over-the-counter dental products. Many of these are intended for topical use only, but some inadvertent ingestion, especially by young children, is unavoiciable. The most effective ap
48 Health Effects of Ingested Fluoride preach to stabilizing the prevalence and severity of dental fluorosis, without jeopardizing the benefits to oral health, is likely to come from more judicious control of fluoride in foods, processed beverages, and dental products, rather than a reduction in the recommended concentra- tions of fluoride in drinking water. But applying such a policy would be formidable; reduction of fluoride concentrations in drinking water would be easier to administer, monitor, and evaluate. If it proved to be scientif- ically justified, such a policy could be considered. Current regulations of EPA are that the maximum contaminant level (MC L) for fluoride in drinking water is 4.0 mg/L, regardless of mean temperature. That standard is considered low enough to prevent crippling skeletal fluorosis, and dental fluorosis is accepted as a purely cosmetic defect with no general health ramifications. However, the most severe forms of dental fluorosis might be more than a cosmetic defect if enough fluorotic enamel is fractured and lost to cause pain, adversely affect food choices, compromise chewing efficiency, and require complex dental treatment. Severe dental fluorosis has been seen in the United States at 3.5 mg/I~ (in a warm climate, 5 times the recommended fluoride con- centration), but even in that community the ramifications of fluorosis were insufficient to recommend a reduction in the MCL. This conclusion points out the need to revise the PHS guidelines related to temperature and for more information on the impact on general health that results from damage to teeth as a consequence of the severest forms of fluorosis. If recommended fluoride concentrations in drinking water should still be expressed in a range related to mean temperature, the MCL might also be expressed more logically as a temperature-related range rather than a single figure. Overall, the evidence of the current current relation of dental fluorosis to fluoride in drinking water in the United States is still sparse, and the evidence that does exist is too inconsistent to be used as a basis for recommending changes in EPA regulations. When the results of further research become available, EPA's regulations might need further review and modification. RESEARCH RECOMMENDATIONS Studies should be conducted on the sources of fluoride during the
Dental Fluorosis 49 critical stages of tooth development in children and on the contribution of the various sources to dental fluorosis etiology. Such information would permit more precise regulation of fluoride products to control fluorosis while retaining fluoride's substantial cariostatic benefits. Studies should be conducted on the relation between water fluoride concentrations and dental fluorosis in various climatic zones. Findings could serve as a basis for any needed revision of the 1962 PHS guide- lines. The lowest concentration of fluoride in toothpaste that produces ac- ceptable cariostasis should be determined. That information would permit the marketing of children's toothpastes that would retain anticaries benefits while minimizing the risk of fluorosis. Further studies should be conducted on the contribution of ingested fluoride and fluoride applied topically to teeth to prevent caries. The results would permit more efficient use of fluoride for caries prevention, thus reducing the risk of fluorosis.