2
Measures of Exposure to Fluoride in the United States

The major sources of internal exposure of individuals to fluorides are the diet (food, water, beverages) and fluoride-containing dental products (toothpaste, fluoride supplements). Internal exposure to fluorides also can occur from inhalation (cigarette smoke, industrial emissions), dermal absorption (from chemicals or pharmaceuticals), ingestion or parenteral administration of fluoride-containing drugs, and ingestion of fluoride-containing soil. Information on the pharmacokinetics of fluoride are provided in Chapter 3.

The National Research Council’s (NRC’s) 1993 review of the health effects of ingested fluoride reported estimates of average daily fluoride intake from the diet of 0.04-0.07 milligrams per kilogram (mg/kg) of body weight for young children in an area with fluoridated water (fluoride concentration in drinking water, 0.7-1.2 mg per liter [L]; NRC 1993). Dietary intake of fluoride by adults in an area with fluoridated water was variously estimated to be between 1.2 and 2.2 mg/day (0.02-0.03 mg/kg for a 70-kg adult). The fluoride intake from toothpaste or mouth rinse by children with good control of swallowing, assuming twice-a-day use, was estimated to equal the intake from food, water, and beverages. The review acknowledged that “substantially” higher intakes of fluoride from consumption of fluoridated water would result for individuals such as outdoor laborers in warm climates or people with high-urine-output disorders, but these intakes were not quantified. Similarly, children and others with poor control of swallowing could have intakes of fluoride from dental products that exceed the dietary intakes, but these intakes also were not quantified. Other factors cited as affecting individual fluoride intakes include changes in the guidelines for



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 23
Fluoride in Drinking Water: A Scientific Review of EPA’S Standards 2 Measures of Exposure to Fluoride in the United States The major sources of internal exposure of individuals to fluorides are the diet (food, water, beverages) and fluoride-containing dental products (toothpaste, fluoride supplements). Internal exposure to fluorides also can occur from inhalation (cigarette smoke, industrial emissions), dermal absorption (from chemicals or pharmaceuticals), ingestion or parenteral administration of fluoride-containing drugs, and ingestion of fluoride-containing soil. Information on the pharmacokinetics of fluoride are provided in Chapter 3. The National Research Council’s (NRC’s) 1993 review of the health effects of ingested fluoride reported estimates of average daily fluoride intake from the diet of 0.04-0.07 milligrams per kilogram (mg/kg) of body weight for young children in an area with fluoridated water (fluoride concentration in drinking water, 0.7-1.2 mg per liter [L]; NRC 1993). Dietary intake of fluoride by adults in an area with fluoridated water was variously estimated to be between 1.2 and 2.2 mg/day (0.02-0.03 mg/kg for a 70-kg adult). The fluoride intake from toothpaste or mouth rinse by children with good control of swallowing, assuming twice-a-day use, was estimated to equal the intake from food, water, and beverages. The review acknowledged that “substantially” higher intakes of fluoride from consumption of fluoridated water would result for individuals such as outdoor laborers in warm climates or people with high-urine-output disorders, but these intakes were not quantified. Similarly, children and others with poor control of swallowing could have intakes of fluoride from dental products that exceed the dietary intakes, but these intakes also were not quantified. Other factors cited as affecting individual fluoride intakes include changes in the guidelines for

OCR for page 23
Fluoride in Drinking Water: A Scientific Review of EPA’S Standards fluoride supplementation and use of bottled water or home water purification systems rather than fluoridated municipal water. The NRC (1993) recommended further research to “determine and compare the intake of fluoride from all sources, including fluoride-containing dental products, in fluoridated and nonfluoridated communities.” This chapter provides a review of the available information on fluoride exposures in the United States, including sources of fluoride exposure, intakes from various fluoride sources, and factors that could affect individual exposures to fluorides. Population subgroups with especially high exposures are discussed. The major emphasis of this chapter is on chronic exposure rather than acute exposure. The use of biomarkers as alternative approaches to estimation of actual individual exposures is also discussed. In practice, most fluorine added to drinking water is in the form of fluosilicic acid (fluorosilicic acid, H2SiF6) or the sodium salt (sodium fluosilicate, Na2SiF6), collectively referred to as fluorosilicates (CDC 1993); for some smaller water systems, fluoride is added as sodium fluoride (NaF). Fluoride in toothpaste and other dental products is usually present as sodium fluoride (NaF), stannous fluoride (SnF2), or disodium monofluorophosphate (Na2PO3F). Fluorine-containing pesticides and pharmaceuticals also contribute to total fluorine exposures and are considered separately. Fluoride in food and drinking water usually is considered in terms of total fluorine content, assumed to be present entirely as fluoride ion (F−). Information on exposures to fluorosilicates and aluminofluorides is also included. SOURCES OF FLUORIDE EXPOSURE Drinking Water General Population The major dietary source of fluoride for most people in the United States is fluoridated municipal (community) drinking water, including water consumed directly, food and beverages prepared at home or in restaurants from municipal drinking water, and commercial beverages and processed foods originating from fluoridated municipalities. On a mean per capita basis, community (public or municipal) water constitutes 75% of the total water ingested in the United States; bottled water constitutes 13%, and other sources (e.g., wells and cisterns) constitute 10% (EPA 2000a). Municipal water sources that are not considered “fluoridated” could contain low concentrations of naturally occurring fluoride, as could bottled water and private wells, depending on the sources. An estimated 162 million people in the United States (65.8% of the population served by public water systems) received “optimally fluori-

OCR for page 23
Fluoride in Drinking Water: A Scientific Review of EPA’S Standards dated”1 water in 2000 (CDC 2002a). This represents an increase from 144 million (62.1%) in 1992. The total number of people served by public water systems in the United States is estimated to be 246 million; an estimated 35 million people obtain water from other sources such as private wells (CDC 2002a,b). The U.S. Environmental Protection Agency (EPA) limits the fluoride that can be present in public drinking-water supplies to 4 mg/L (maximum contaminant level, or MCL) to protect against crippling skeletal fluorosis, with a secondary maximum contaminant level (SMCL) of 2 mg/L to protect against objectionable enamel fluorosis (40CFR 141.62(b)[2001], 40CFR 143.3[2001]). Of the 144 million people with fluoridated public water supplies in 1992, approximately 10 million (7%) received naturally fluoridated water, the rest had artificially fluoridated water (CDC 2002c). Of the population with artificially fluoridated water in 1992, more than two-thirds had a water fluoride concentration of 1.0 mg/L, with almost one-quarter having lower concentrations and about 5% having concentrations up to 1.2 mg/L (CDC 1993; see Appendix B). Of the approximately 10 million people with naturally fluoridated public water supplies in 1992, approximately 67% had fluoride concentrations ≤ 1.2 mg/L (CDC 1993; see Appendix B). Approximately 14% had fluoride concentrations between 1.3 and 1.9 mg/L and another 14% had between 2.0 and 3.9 mg/L; 2% (just over 200,000 persons) had natural fluoride concentrations equal to or exceeding 4.0 mg/L.2 Water supplies that exceeded 4.0 mg/L ranged as high as 11.2 mg/L in Colorado, 12.0 mg/L in Oklahoma, 13.0 mg/L in New Mexico, and 15.9 mg/L in Idaho (see Appendix B, Table B-3).3 States with the largest populations receiving water supplies with fluoride at ≥ 4.0 mg/L included Virginia (18,726 persons, up to 6.3 mg/L), Oklahoma (18,895 persons, up to 12.0 mg/L), Texas (36,863 persons, up to 8.8 mg/L), and South Carolina (105,618 persons, up to 5.9 mg/L). Little information is available on the fluoride content of private water sources, but the variability can reasonably be expected to be high and to 1 The term optimally fluoridated water means a fluoride level of 0.7-1.2 mg/L; water fluoride levels are based on the average maximum daily air temperature of the area (see Appendix B). 2 More recently (2000), CDC has estimated that 850,000 people are served by public water supplies containing fluoride in excess of 2 mg/L; of these, 152,000 people receive water containing fluoride in excess of 4 mg/L (unpublished data from CDC as reported in EPA 2003a). Based on analytical data from 16 states, EPA (2003a) estimates that 1.5-3.3 million people nationally are served by public water supplies with fluoride concentrations exceeding 2 mg/L; of these 118,000-301,000 people receive water with fluoride concentrations greater than 4 mg/L. 3 High-fluoride municipal waters are generally found in regions that have high fluoride concentrations in the groundwater or in surface waters. ATSDR (2003) has reviewed fluoride concentrations in environmental media, including groundwater and surface water. Fleischer (1962) and Fleischer et al. (1974) reported fluoride concentrations in groundwater by county for the coterminous United States.

OCR for page 23
Fluoride in Drinking Water: A Scientific Review of EPA’S Standards depend on the region of the country. Fluoride measured in well water in one study in Iowa ranged from 0.06 to 7.22 mg/L (mean, 0.45 mg/L); home-filtered well water contained 0.02-1.00 mg/L (mean, 0.32 mg/L; Van Winkle et al. 1995). Hudak (1999) determined median fluoride concentrations for 237 of 254 Texas counties (values were not determined for counties with fewer than five observations). Of the 237 counties, 84 have median groundwater fluoride concentrations exceeding 1 mg/L; of these, 25 counties exceed 2 mg/L and five exceed 4 mg/L. Residents in these areas (or similar areas in other states) who use groundwater from private wells are likely to exceed current guidelines for fluoride intake. Duperon et al. (1995) pointed out that fluoride concentrations reported by local water suppliers can be substantially different from concentrations measured in water samples obtained in homes. Use of home water filtration or purification systems can reduce the fluoride concentration in community water by 13% to 99%, depending on the type of system (Duperon et al. 1995; Van Winkle et al. 1995; Jobson et al. 2000). Distillation or reverse osmosis can remove nearly all the fluoride. The extent of use of home water filtration or purification systems nationally is not known but obviously would affect the fluoride intake for people using such systems. Van Winkle et al. (1995) reported that 11% of their study population (in Iowa) used some type of home filtration either for well water or for public water. Fluoride concentrations in bottled water4 are regulated by law to a maximum of 1.4-2.4 mg/L if no fluoride is added and a maximum of 0.8-1.7 mg/L if fluoride is added (the applicable value within the range depends on the annual average of maximum daily air temperatures at the location of retail sale; 21CFR 165.110[2003]). Maximum fluoride concentrations for imported bottled water are 1.4 mg/L if no fluoride is added and 0.8 mg/L if fluoride is added (21CFR 165.110[2003]). Fluoride concentrations are required on labels in the United States only if fluoride is added. Fluoride concentrations listed on labels or in chemical analyses available on the Internet for various brands range from 0 to 3.6 mg/L (Bartels et al. 2000; Johnson and DeBiase 2003; Bottled Water Web 2004); of those without added fluoride, most are below 0.6 mg/L. Most brands appear to list fluoride content only if they are specifically advertising the fact that their water is fluoridated; fluoride concentrations of these brands range from 0.5 to 0.8 mg/L (for “nursery” or “infant” water) up to 1.0 mg/L. Several reports indicate 4 The term “bottled water” applies to water intended for human consumption, containing no added ingredients besides fluoride or appropriate antimicrobial agents; the regulations apply to bottled water, drinking water, artesian water, artesian well water, groundwater, mineral water, purified water, demineralized water, deionized water, distilled water, reverse osmosis water, purified drinking water, demineralized drinking water, deionized drinking water, distilled drinking water, reverse osmosis drinking water, sparkling water, spring water, and well water (21CFR 165.110[2003]).

OCR for page 23
Fluoride in Drinking Water: A Scientific Review of EPA’S Standards that fluoride concentrations obtained from the manufacturer or stated on labels for bottled waters might not be accurate (Weinberger 1991; Toumba et al. 1994; Bartels et al. 2000; Lalumandier and Ayers 2000; Johnson and DeBiase 2003; Zohouri et al. 2003). Measured fluoride concentrations in bottled water sold in the United States have varied from 0 to 1.36 mg/L (Nowak and Nowak 1989; Chan et al. 1990; Stannard et al. 1990; Van Winkle et al. 1995; Bartels et al. 2000; Lalumandier and Ayers 2000; Johnson and DeBiase 2003). Van Winkle et al. (1995) reported a mean of 0.18 mg/L for 78 commercial bottled waters in Iowa. Johnson and DeBiase (2003) more recently reported values ranging from 0 to 1.2 mg/L for 65 bottled waters purchased in West Virginia, with 57 brands having values below 0.6 mg/L. Measured fluoride concentrations in bottled waters in other countries have similar ranges: 0.05-4.8 mg/L in Canada (Weinberger 1991), 0.10-0.80 mg/L in the United Kingdom (Toumba et al. 1994), and 0.01-0.37 mg/L more recently in the United Kingdom (Zohouri et al. 2003).5 Bartels et al. (2000) found significant variation in fluoride concentrations among samples of the same brand with different bottling dates purchased in the same city. In general, distilled and purified (reverse osmosis) waters contain very low concentrations of fluoride; drinking water (often from a municipal tap) and spring water vary with their source, as do mineral waters, which can be very low or very high in fluoride. Most spring water sold in the United States probably has a low fluoride content (<0.3 mg/L). Typical fluoride concentrations in various types of drinking water in the United States are summarized in Table 2-1. Average per capita ingestion of community or municipal water is estimated to be 927 mL/day (EPA 2000a; see Appendix B6). The estimated 90th percentile of the per capita ingestion of community water from that survey is 2.016 L/day. Estimated intakes by those actually consuming community water (excluding people with zero ingestion of community water) are higher, with a mean of 1.0 L/day and a 90th percentile of 2.069 L/day (EPA 2000a). Thus, if national estimates of water intake (see Appendix B) 5 The European Commission has set a maximum limit of 5.0 mg/L for fluoride in natural mineral waters, effective January 1, 2008 (EC 2003). In addition, natural mineral waters with a fluoride concentration exceeding 1.5 mg/L must be labeled with the words “contains more than 1.5 mg/L of fluoride: not suitable for regular consumption by infants and children under 7 years of age,” and for all natural mineral waters, the actual fluoride content is to be listed on the label. England has essentially the same requirements (TSO 2004), applicable to all bottled waters (natural mineral waters, spring water, and bottled drinking water). 6 As described more fully in Appendix B, the values from EPA (2000a) are from a short-term survey of more than 15,000 individuals in the United States. Although these values are considered reasonable indicators both of typical water consumption and of the likely range of water consumption on a long-term basis, they should not be used by themselves to predict the number of individuals or percentage of the population that consumes a given amount of water on a long-term basis.

OCR for page 23
Fluoride in Drinking Water: A Scientific Review of EPA’S Standards TABLE 2-1 Typical Fluoride Concentrations of Major Types of Drinking Water in the United States Source Range, mg/La Municipal water (fluoridated) 0.7-1.2 Municipal water (naturally fluoridated) 0.7-4.0+ Municipal water (nonfluoridated) <0.7 Well water 0-7+ Bottled water from municipal source 0-1.2 Spring water 0-1.4 (usually <0.3) Bottled “infant” or “nursery” water 0.5-0.8 Bottled water with added fluorideb 0.8-1.0 Distilled or purified water <0.15 aSee text for relevant references. bOther than “infant” or “nursery” water. are assumed to be valid for the part of the population with fluoridated water supplies, the intake of fluoride for a person with average consumption of community water (1 L/day) in a fluoridated area ranges from 0.7 to 1.2 mg/day, depending on the area. A person with consumption of community water equivalent to the 90th percentile in that survey (2.069 L/day) would have a fluoride intake between 1.4 and 2.5 mg/day, from community water alone. Table 2-2 provides examples of fluoride intake by typical and high consumers of municipal water by age group. The estimates of water consumption described in Appendix B are in keeping with recently published “adequate intake” values for total water consumption (including drinking water, all beverages, and moisture in food; IOM 2004; see Appendix B, Table B-10). Note that these estimates are national values; the range of values for optimal fluoridation was intended to account for expected regional differences in water consumption due to regional temperature differences (see Appendix B). A separate study based on the same data used by EPA (2000a) found no strong or consistent association between water intake and month or season (Heller et al. 1999). Another recent study of American children aged 1-10 years also found no significant relationship between water consumption and mean temperature in modern conditions (perhaps due to artificial temperature regulation) and suggested that the temperature-related guidelines for fluoride concentrations in drinking water be reevaluated (Sohn et al. 2001). Actual intakes of fluoride from drinking water by individuals depend on their individual water intakes, the source or sources of that water, and the use of home water purification or filtration systems. As described earlier, fluoride concentrations in community water might vary from their reported concentrations; fluoride content of bottled water also varies considerably with brand or source, with packaging date for a given brand, and from

OCR for page 23
Fluoride in Drinking Water: A Scientific Review of EPA’S Standards TABLE 2-2 Examples of Fluoride Intake from Consumption of Community (Municipal) Water by People Living in Fluoridated Areasa   Typical Consumersb High Consumersc   Water Consumption Fluoride Intaked Water Consumption Fluoride Intaked   mL/day mL/kg/day mg/day mg/kg/day mL/day mL/kg/day mg/day mg/kg/day U.S. population (total) 1,000 17 0.7-1.2 0.012-0.020 2,100 33 1.5-2.5 0.023-0.040 All infants (<1 year)e 500 60 0.35-0.6 0.042-0.072 950 120 0.67-1.1 0.084-0.14 Children 1-2 years 350 26 0.25-0.42 0.018-0.031 700 53 0.49-0.84 0.037-0.064 Children 3-5 years 450 23 0.32-0.54 0.016-0.028 940 45 0.66-1.1 0.032-0.054 Children 6-12 years 500 16 0.35-0.6 0.011-0.019 1,000 33 0.7-1.2 0.023-0.040 Youths 13-19 years 800 12 0.56-0.96 0.0084-0.014 1,700 26 1.2-2.0 0.018-0.031 Adults 20-49 years 1,100 16 0.77-1.3 0.011-0.019 2,200 32 1.5-2.6 0.022-0.038 Adults 50+ years 1,200 17 0.84-1.4 0.012-0.020 2,300 32 1.6-2.8 0.022-0.038 Females 13-49 yearsf 980 15 0.69-1.2 0.011-0.018 2,050 32 1.4-2.5 0.022-0.038 aBased on consumption data described in Appendix B for people actually consuming community (municipal) water. bBased on a typical consumption rate of community (municipal) water for the age group. cBased on a reasonably high (but not upper bound) consumption rate of community (municipal) water for the age group; some individual exposures could be higher. dBased on fluoride concentrations of 0.7-1.2 mg/L. eIncludes any infant, nursing or nonnursing, who consumes at least some community water; these infants may be fed primarily breast milk, ready-to-feed formula (to which no water is normally added), or formula prepared from concentrate (which requires addition of water). fWomen of childbearing age.

OCR for page 23
Fluoride in Drinking Water: A Scientific Review of EPA’S Standards information (if any) given on the labels or provided by the manufacturer. Private water sources (e.g., wells and cisterns) probably are even more variable in fluoride content, with some regions of the country being especially high and others very low. A number of authors have pointed out the difficulty doctors and dentists face in ascertaining individual fluoride intakes, just from drinking water (from all sources), for the purpose of prescribing appropriate fluoride supplementation (Nowak and Nowak 1989; Chan et al. 1990; Stannard et al. 1990; Levy and Shavlick 1991; Weinberger 1991; Dillenberg et al. 1992; Jones and Berg 1992; Levy and Muchow 1992; Toumba et al. 1994; Duperon et al. 1995; Van Winkle et al. 1995; Heller et al. 1999; Bartels et al. 2000; Lalumandier and Ayers 2000; Johnson and DeBiase 2003; Zohouri et al. 2003). High Intake Population Subgroups EPA, in its report to Congress on sensitive subpopulations (EPA 2000b), defines sensitive subpopulations in terms of either their response (more severe response or a response to a lower dose) or their exposure (greater exposure than the general population). Hence, it is appropriate to consider those population subgroups whose water intake is likely to be substantially above the national average for the corresponding sex and age group. These subgroups include people with high activity levels (e.g., athletes, workers with physically demanding duties, military personnel); people living in very hot or dry climates, especially outdoor workers; pregnant or lactating women; and people with health conditions that affect water intake. Such health conditions include diabetes mellitus, especially if untreated or poorly controlled; disorders of water and sodium metabolism, such as diabetes insipidus; renal problems resulting in reduced clearance of fluoride; and short-term conditions requiring rapid rehydration, such as gastrointestinal upsets or food poisoning (EPA 2000a). (While the population sample described in Appendix B [Water Ingestion and Fluoride Intakes] included some of these individuals, the study did not attempt to estimate means or distributions of intake for these specific subgroups.) As shown in Appendix B (Tables B-4 to B-9), some members of the U.S. population could have intakes from community water sources of as much as 4.5-5 L/day (as high as 80 mL/kg/day for adults). Some infants have intakes of community water exceeding 200 mL/kg/day. Heller et al. (1999), using the same data set as EPA (2000a), reported that 21 of 14,640 people (of all ages) had water intakes over 6 standard deviations from the mean (greater than 249 mL/kg/day). Whyte et al. (2005) describe an adult woman who consistently consumed 1-2 gallons (3.8-7.6 L) of fluid per day (instant tea made with well water); no specific reason for her high fluid consumption is given.

OCR for page 23
Fluoride in Drinking Water: A Scientific Review of EPA’S Standards Fluid requirements of athletes, workers, and military personnel depend on the nature and intensity of the activity, the duration of the activity, and the ambient temperature and humidity. Total sweat losses for athletes in various sports can range from 200 to 300 mL/hour to 2,000 mL/hour or more (Convertino et al. 1996; Horswill 1998; Cox et al. 2002; Coyle 2004). Most recommendations on fluid consumption for athletes are concerned with matching fluid replacement to fluid losses during the training session or competition to minimize the detrimental effects of dehydration on athletic performance (Convertino et al. 1996; Horswill 1998; Coris et al. 2004; Coyle 2004). Depending on the nature of the sport or training session, the ease of providing fluid, and the comfort of the athlete with respect to content of the gastrointestinal tract, fluid intake during exercise is often only a fraction (e.g., one-half) of the volume lost, and losses of 2% of body weight or more might occur during an exercise session in spite of fluid consumption during the session (Convertino et al. 1996; Cox et al. 2002; Coris et al. 2004; Coyle 2004). Total daily fluid consumption by athletes generally is not reported; for many athletes, it is probably on the order of 5% of body weight (50 mL/ kg/day) or more to compensate for urinary and respiratory losses as well as sweat losses. For example, Crossman (2003) described a professionally prepared diet plan for a major league baseball player that includes 26 cups (6.2 L) of water or sports drink on a workout day and 19 cups (4.5 L) on an off-day; this is in addition to 9-11 cups (2.1-2.6 L) of milk, fruit juice, and sports drink with meals and scheduled snacks (total fluid intake of 6.8-8.8 L/day, or 52-67 mL/kg/day for a 132-kg player7). While some players and teams probably use bottled or distilled water, most (especially at the amateur and interscholastic levels) probably use local tap water; also, sports drinks might be prepared (commercially or by individuals) with tap water. The U.S. Army’s policy on fluid replacement for warm-weather training calls for 0.5-1 quart/hour (0.47-0.95 L/hour), depending on the temperature, humidity, and type of work (Kolka et al. 2003; USASMA 2003). In addition, fluid intake is not to exceed 1.5 quarts/hour (1.4 liter/hour) or 12 quarts/day (11.4 L/day). The Army’s planning factor for individual tap water consumption ranges from 1.5 gallons/day (5.7 L/day) for temperate conditions to 3.0 gallons/day (11.4 L/day) for hot conditions (U.S. Army 1983). Hourly intake can range from 0.21 to 0.65 L depending on the temperature (McNall and Schlegel 1968), and daily intake among physically active individuals can range from 6 to 11 L (U.S. Army 1983, cited by EPA 1997). Nonmilitary outdoor workers in hot or dry climates probably would have similar needs. 7 The player’s weight was obtained from the 2003 roster of the Cleveland Indians baseball team (http://cleveland.indians.mlb.com).

OCR for page 23
Fluoride in Drinking Water: A Scientific Review of EPA’S Standards Water intakes for pregnant and lactating women are listed separately in Appendix B (Tables B-4 to B-9). Total water intake for pregnant women does not differ greatly from that for all adult females (Table B-9), while total water consumption by lactating women is generally higher. For the highest consumers among lactating women, consumption rates approximate those for athletes and workers (50-70 mL/kg/day). Diabetes mellitus and diabetes insipidus are both characterized by high water intakes and urine volumes, among other things (Beers and Berkow 1999; Eisenbarth et al. 2002; Robinson and Verbalis 2002; Belchetz and Hammond 2003). People with untreated or poorly controlled diabetes mellitus would be expected to have substantially higher fluid intakes than nondiabetic members of the population. The American Diabetes Association (2004) estimates that 18.2 million people in the United States (6.3% of the population) have diabetes mellitus and that 5.2 million of these are not aware they have the disease. Other estimates range from 16 to 20 million people in the United States, with up to 50% undiagnosed (Brownlee et al. 2002; Buse et al. 2002). Diabetes insipidus, or polyuria, is defined as passage of large volumes of urine, in excess of about 2 L/m2/day (approximately 150 mL/kg/day at birth, 110 mL/kg/day at 2 years, and 40 mL/kg/day in older children and adults) (Baylis and Cheetham 1998; Cheetham and Baylis 2002). Diabetes insipidus includes several types of disease distinguished by cause, including both familial and acquired disorders (Baylis and Cheetham 1998; Cheetham and Baylis 2002; Robinson and Verbalis 2002). Water is considered a therapeutic agent for diabetes insipidus (Beers and Berkow 1999; Robinson and Verbalis 2002); in addition, some kinds of diabetes insipidus can be treated by addressing an underlying cause or by administering vasopressin (antidiuretic hormone) or other agents to reduce polyuria to a tolerable level. The Diabetes Insipidus Foundation (2004) estimates the number of diabetes insipidus patients in the United States at between 40,000 and 80,000. Someone initially presenting with central or vasopressin-sensitive diabetes insipidus might ingest “enormous” quantities of fluid and may produce 3-30 L of very dilute urine per day (Beers and Berkow 1999) or up to 400 mL/kg/day (Baylis and Cheetham 1998). Most patients with central diabetes insipidus have urine volumes of 6-12 L/day (Robinson and Verbalis 2002). Patients with primary polydipsia might ingest and excrete up to 6 L of fluid per day (Beers and Berkow 1999). Pivonello et al. (1998) listed water intakes of 5.5-8.6 L/day for six adults with diabetes insipidus who did not take vasopressin and 1.4-2.5 L/day for 12 adults who used a vasopressin analogue. An estimated 20% to 40% of patients on lithium therapy have a urine volume > 2.5 L/day, and up to 12% have frank nephrogenic diabetes insipidus characterized by a urine volume > 3 L/day (Mukhopadhyay et al. 2001).

OCR for page 23
Fluoride in Drinking Water: A Scientific Review of EPA’S Standards Five papers described enamel fluorosis in association with diabetes insipidus or polydipsia (Table 2-3). Two of the papers described cases of enamel fluorosis in the United States resulting from fluoride concentrations of 1, 1.7, or 2.6 mg/L in drinking water (Juncos and Donadio 1972; Greenberg et al. 1974). The two individuals drinking water with fluoride at 1.7 and 2.6 mg/L also had roentgenographic bone changes consistent with “systemic fluorosis”8 (Juncos and Donadio 1972). These patients and four other renal patients in the U.S. “in whom fluoride may have been the cause of detectable clinical and roentgenographic effects” were also reported by Johnson et al. (1979); most of the patients had urine volumes exceeding 3 L/day and drinking water with fluoride concentrations around 1.7-3 mg/L. Moderate and severe enamel fluorosis have been reported in diabetes insipidus patients in other countries with drinking water containing fluoride at 0.5 mg/L (Klein 1975) or 1 mg/L (Seow and Thomsett 1994), and severe enamel fluorosis with skeletal fluorosis has been reported with fluoride at 3.4 mg/L (Mehta et al. 1998). Greenberg et al. (1974) recommended that children with any disorder that gives rise to polydipsia and polyuria9 be supplied a portion of their water from a nonfluoridated source. Table 2-4 provides examples of fluoride intake by members of several population subgroups characterized by above-average water consumption (athletes and workers, patients with diabetes mellitus or diabetes insipidus). It should be recognized that, for some groups of people with high water intakes (e.g., those with a disease condition or those playing indoor sports such as basketball or hockey), there probably will be little correlation of water intake with outdoor temperature—such individuals in northern states would consume approximately the same amounts of water as their counterparts in southern states. However, fluoridation still varies from state to state (Appendix B), so that some individuals could consume up to 1.7 times as much as others for the same water intake (1.2 versus 0.7 mg/L). Background Food Measured fluoride in samples of human breast milk is very low. Dabeka et al. (1986) found detectable concentrations in only 92 of 210 samples (44%) obtained in Canada, with fluoride ranging from <0.004 to 0.097 mg/L. The mean concentration in milk from mothers in fluoridated 8 These two individuals also had impaired renal function, which could have increased their retention of fluoride (see Chapter 3). 9 Greenberg et al. (1974) listed “central diabetes insipidus, psychogenic water ingestion, renal medullary disease, including hypercalemic nephropathy, hypokalemic nephropathy and anatomic and vascular disturbances and those diseases causing solute diuresis” as disorders associated with “excessive” consumption of water and therefore the possibility of “fluoride toxicity in a community with acceptable fluoride concentration.”

OCR for page 23
Fluoride in Drinking Water: A Scientific Review of EPA’S Standards than in toenails in 2- to 6-year old children and showed a correlation between nail concentrations and dietary fluoride intake (exclusive of fluoride in toothpaste). Plasma fluoride in these children was not correlated with fluoride in fingernails, toenails, diet, or drinking water. In contrast, Corrêa Rodrigues et al. (2004), in samples from 2- to 3-year-old children, found no significant differences in fluoride concentrations between fingernails and toenails collected at the same time. An increase in fluoride intake in these children was reflected in nail samples approximately 4 months later (Corrêa Rodrigues et al. 2004). Most likely, differences in “lag times” and differences between fingernails and toenails in the same individual reflect differences in growth rates of the nails due to factors such as age or differences in blood flow. McDonnell et al. (2004) found a wide variation in growth rates of thumbnails of 2- and 3-year-old children; age, gender, and fluoride exposure had no effect on the growth rates. However, it was emphasized that, for any study in which it is of interest to estimate the timing of a fluoride exposure based on measurements of fluoride in nails, the growth rate of the nails should be measured for each individual. Czarnowski et al. (1999) found correlations between water fluoride concentrations and urinary fluoride, fluoride in hair, and bone mineral density measured in 300 people in the Gdánsk region of Poland. For workers with occupational exposure to airborne fluoride (largely HF), Czarnowski and Krechniak (1990) found good correlation among groups of workers between fluoride concentrations in urine and nails (r = 0.99); correlation between concentrations in urine and hair or hair and nails was also positive but not as good (r = 0.77 and 0.70, respectively). For individual values, positive correlation was found only between concentrations in urine and nails (r = 0.73). It was not possible to establish correlations between fluoride concentrations in biological media and air (Czarnowski and Krechniak 1990). Measuring the fluoride content of teeth and bones can give an indication of chronic or cumulative fluoride exposure, although after cessation of fluoride exposure, bone fluoride concentrations slowly decrease because of resorption of bone. In addition, bone turnover results in the accumulation of various concentrations of fluoride in different bone types and sites (Selwitz 1994). Dentin has also been suggested as a reasonably accurate marker for long-term exposure (Selwitz 1994), although Vieira et al. (2005) found no correlation between bone fluoride and either enamel or dentin fluoride in persons with exposure to 0.07 or 1.0 mg/L fluoride in drinking water. Roholm (1937) reported that the fluoride content in normal teeth varied from 190 to 300 ppm (0.19 to 0.30 mg/g) in the total ash, with 5-7 times as much fluoride in the dentin as in the enamel. Fluoride content in the total ash of teeth from five cryolite workers (employed 8-10 years; three with osteosclerosis) contained 1,100-5,300 ppm (1.1-5.3 mg/g), with the most carious teeth containing the most fluoride. Roholm (1937) also reported

OCR for page 23
Fluoride in Drinking Water: A Scientific Review of EPA’S Standards normal bone fluoride concentrations of 480-2,100 ppm in bone ash (0.48-2.1 mg/g bone ash in ribs), with concentrations between 3,100 and 13,100 ppm in bone ash (3.1 and 13.1 mg/g bone ash; varying with type of bone) in two cryolite workers. Hodge and Smith (1965), summarizing several reports, listed mean concentrations of bone fluoride in normal individuals between 450 and 1,200 ppm in bone ash and in people “suffering excessive exposure” to fluorides between 7,500 and 20,830 ppm in bone ash. More recently, Eble et al. (1992) have reported fluoride concentrations in bone ash ranging from 378 ppm (16-year old with <0.2 mg/L fluoride in drinking water since infancy) to 3,708 ppm (79-year old with fluoridated water). A 46-year old female with chronic renal failure had a fluoride concentration in bone ash of 3,253 ppm (Eble et al. 1992). The data of Zipkin et al. (1958) shows a good relationship between drinking-water fluoride and the mean percentage of fluoride in bone (iliac crest, rib, and vertebra) for adults in areas of various fluoride concentrations in drinking water. However, the ranges (Table 2-16; see also Chapter 3, Figure 3-1) suggest that variability among individuals within groups could be large, probably reflecting variability in individual fluoride intakes, duration of exposure, and age. A major disadvantage of measuring bone fluoride is the invasiveness of bone sampling in live individuals. Although easier to do, x-ray screening for increased bone density should be done only when the need for information justifies the radiation dose involved; in addition, bone density might not be related solely to fluoride exposure or to bone fluoride content. The two most important biomarkers of effect for fluoride are considered to be enamel fluorosis and skeletal fluorosis (ATSDR 2003); these are discussed more fully in Chapters 4 and 5. Enamel fluorosis is characterized by mottling and erosion of the enamel of the teeth and is associated with elevated fluoride intakes during the childhood years when the teeth are developing. According to the U.S. Public Health Service (PHS 1991), both the percent prevalence and the increasing severity of enamel fluorosis are associated with increasing fluoride concentration in drinking water (and presumably actual fluoride intake). For “optimally” fluoridated water (0.7-1.2 mg/L), 22% of children examined in the 1980s showed some fluorosis (mostly very mild or mild); at water fluoride concentrations above 2.3 mg/L, more than 70% of children showed fluorosis (PHS 1991; NRC 1993). Some children developed fluorosis even at the lowest fluoride concentrations (<0.4 mg/L), suggesting that either fluoride intakes are variable within a population with the same water supply or there is variability in the susceptibility to fluorosis within populations (or both). Baelum et al. (1987) indicated that 0.03 mg/kg/day might not be protective against enamel fluorosis, and Fejerskov et al. (1987) stated that the borderline dose above which enamel fluorosis might develop could be as low as 0.03 mg/kg/day.

OCR for page 23
Fluoride in Drinking Water: A Scientific Review of EPA’S Standards DenBesten (1994) described the limitations of using enamel fluorosis as a biomarker of exposure: enamel fluorosis is useful only for children less than about 7 years old when the exposure occurred; the incidence and degree of fluorosis vary with the timing, duration, and concentration; and there appear to be variations in individual response. Selwitz (1994), summarizing a workshop on the assessment of fluoride accumulation, also indicated that variability in response (incidence and severity of enamel fluorosis) to fluoride exposure may result from physiological differences among individuals and that enamel fluorosis is not an adequate biomarker for fluoride accumulation or potentially adverse health effects beyond the period of tooth formation. Selwitz (1994) did suggest that enamel fluorosis could be used as a biomarker of fluoride exposure in young children within a community over time. Skeletal fluorosis (see also Chapter 5) is characterized by increased bone mass, increased radiographic density of the bones, and a range of skeletal and joint symptoms; preclinical skeletal fluorosis is associated with fluoride concentrations of 3,500-5,500 ppm in bone ash and clinical stages I, II, and III with concentrations of 6,000-7,000, 7,500-9,000, and >8,400, respectively (PHS 1991), although other sources indicate lower concentrations of bone fluoride in some cases of skeletal fluoride (see Chapter 5). According to the Institute of Medicine, “Most epidemiological research has indicated that an intake of at least 10 mg/day [of fluoride] for 10 or more years is needed to produce clinical signs of the milder forms of [skeletal fluorosis]” (IOM 1997). However, the National Research Council (NRC 1993) indicated that crippling (as opposed to mild) skeletal fluorosis “might occur in people who have ingested 10-20 mg of fluoride per day for 10-20 years.” A previous NRC report (NRC 1977) stated that a retention of 2 mg of fluoride per day (corresponding approximately to a daily intake of 4-5 mg) “would mean that an average individual would experience skeletal fluorosis after 40 yr, based on an accumulation of 10,000 ppm fluoride in bone ash.” Studies in other countries indicate that skeletal fluorosis might be in part a marker of susceptibility as well as exposure, with factors such as dietary calcium deficiency involved in addition to fluoride intake (Pettifor et al. 1989; Teotia et al. 1998). Hodge and Smith (1965) summarized a number of studies of skeletal fluorosis, including two that indicated affected individuals in the United States with water supplies containing fluoride at 4.8 or 8 mg/L. They also stated categorically that “crippling fluorosis has never been seen in the United States.” The individuals with endemic fluorosis at 4.8 mg/L are referred to elsewhere as having “radiographic osteosclerosis, but no evidence of skeletal fluorosis” (PHS 1991). In combination with high fluid intake and large amounts of tea, “the lowest drinking-water concentration of fluoride

OCR for page 23
Fluoride in Drinking Water: A Scientific Review of EPA’S Standards associated with symptomatic skeletal fluorosis that has been reported to date is 3 ppm, outside of countries such as India” (NRC 1977). Both the PHS (1991) and the NRC (1993) indicated that only five cases of crippling skeletal fluorosis have been reported in the literature in the United States (including one case in a recent immigrant from an area with fluoride in the drinking water at 3.9 mg/L) (PHS 1991). These individuals were said to have water supplies ranging from 3.9 to 8.0 mg/L (water fluoride content given for one of the individuals is actually less than 3.9 mg/L) (PHS 1991). Two of the individuals had intakes of up to 6 L/day of water containing fluoride at 2.4-3.5 or 4.0-7.8 mg/L (PHS 1991; NRC 1993); this corresponds to fluoride intakes of up to 14.4-21 or 24-47 mg/day. Several cases of skeletal fluorosis reported in the United States are summarized in Table 2-17. These reports indicate that a fluoride concentration of 7-8 mg/L for 7 years is sufficient to bring about skeletal fluorosis (Felsenfeld and Roberts 1991), but skeletal fluorosis may occur at much lower fluoride concentrations in cases of renal insufficiency (Juncos and Donadio 1972; Johnson et al. 1979). People who consume instant tea are at increased risk of developing skeletal fluorosis, especially if they drink large volumes, use extra-strength preparations, or use fluoridated or fluoride-contaminated water (Whyte et al. 2005). In summary, selecting appropriate biomarkers for a given fluoride study depends on a number of factors, as listed above. A major consideration is the time period of interest for the study (e.g., current or recent exposures versus exposures in childhood versus cumulative exposures) and whether the intent is to demonstrate differences among groups or to characterize exposures of specific individuals. Many of the areas for further research identified by a 1994 workshop (Whitford et al. 1994) are still relevant for improving the assessment of fluoride exposures. FINDINGS Table 2-18 summarizes various published perspectives on the significance of given concentrations of fluoride exposure. Historically, a daily intake of 4-5 mg by an adult (0.057-0.071 mg/kg for a 70-kg adult) was considered a “health hazard” (McClure et al. 1945, cited by Singer et al. 1985). However, the Institute of Medicine (IOM 1997) now lists 10 mg/day as a “tolerable upper intake” for children > 8 years old and adults, although that intake has also been associated with the possibility of mild (IOM 1997) or even crippling (NRC 1993) skeletal fluorosis. The recommended optimal fluoride intake for children to maximize caries prevention and minimize the occurrence of enamel fluorosis is often stated as being 0.05-0.07 mg/kg/day (Levy 1994; Heller et al. 1999, 2000). Burt (1992) attempted to track down the origin of the estimate of 0.05-0.07

OCR for page 23
Fluoride in Drinking Water: A Scientific Review of EPA’S Standards TABLE 2-17 Case Reports of Skeletal Fluorosis in the United States Study Subjects Exposure Conditions Comments Reference (a) 18-year-old boy, 57.4 kg (b) 17-year-old girl, 45.65 kg (a) “high” intake of well water containing fluoride at 2.6 mg/L since early childhood; current intake, 7.6 L/day (0.34 mg/kg/day) (b) “high” intake of water containing fluoride at 1.7 mg/L since infancy; current intake, 4 L/day (0.15 mg/kg/day) Enamel fluorosis and roentgenographic bone changes consistent with “systemic fluorosis,” attributed to the combination of renal insufficiency and polydipsia (the latter resulting from the renal disease); reported by the Mayo Clinic Juncos and Donadio 1972 Six renal patients seen at the Mayo Clinic over a several year period (includes the two patients reported by Juncos and Donadio) Drinking water with 1.7-3 mg/L fluoride; water consumption not stated, but urine volumes of “most” of the patients exceeded 3 L/day Fluoride “may have been the cause of detectable clinical and roentgenographic effects” Five of the patients had renal disease of at least 15 years duration before skeletal symptoms developed Johnson et al. 1979 54-year-old woman in Oklahoma Well water with fluoride concentration of 7.3-8.2mg/L (382-429 µmol/L); duration of residence at that location, 7 years; prior to that she had used municipal water at less than 2 mg/L fluoride; water consumption not reported, but considered likely to be “increased” due to hot summers Osteosclerosis, elevated serum alkaline phosphatase, stiffness of knees and hips (2 years duration), kyphosis Renal insufficiency was not a factor Felsenfeld and Roberts 1991 52-year-old woman in Missouri Daily consumption of 1-2 gallons (3.8-7.6 L) per day of double-strength instant tea made with unfiltered well water (2.8 mg/L fluoride in the well water) for close to 10 years; estimated fluoride intake of 37-74 mg/day (11-22 mg/day from well water and 26-52 mg/day from tea) Osteosclerosis, increased bone mineral density, bone and joint pains Intake of fluoride from well water alone was considered sufficient to cause mild skeletal fluorosis No mention of any renal disease Whyte et al. 2005

OCR for page 23
Fluoride in Drinking Water: A Scientific Review of EPA’S Standards TABLE 2-18 Summary of Current and Historical Perspectives on Fluoride Exposure Exposure, mg/kg/day Description Reference 0.0014 “Adequate intake” for children < 6 months olda (0.01 mg/day) IOM 1997; ADA 2005 0.01-0.04 Average daily dietary fluoride intake for children 0-2 years old residing in nonfluoridated areas (< 0.4 mg/L) IOM 1997b 0.017-0.031 Average daily intake by adults in a fluoridated area (1.2-2.2 mg/day)c NRC 1993 0.017-0.054 Lower end of “safe and adequate daily dietary intake” for children ≥ 0-10 yearsd (0.1-1.5 mg/day) NRC 1989b 0.019-0.033 Lower end of “safe and adequate daily dietary intake” for children 10 years and adultsd (1.5 mg/day) NRC 1989b 0.02-0.10 Average daily dietary fluoride intake for children 1-9 years residing in fluoridated areas (0.7-1.1 mg/L) McClure 1943e 0.038-0.069 Upper end of “safe and adequate daily dietary intake” for children ≥ 10 years and adultsd (2.5-4.0 mg/day) NRC 1989b 0.04-0.07 Average daily intake by children in a fluoridated area NRC 1993 0.05 “Adequate intake” for all ages above 6 months olda,f IOM 1997; ADA 2005 0.05 ATSDR’s minimal risk levelg (chronic duration, based on increased rate of bone fractures)h ATSDR 2003 0.05-0.13 Average daily dietary fluoride intake for children 0-2 years old residing in fluoridated areas (0.7-1.1 mg/L) IOM 1997b 0.05-0.07 “Optimal” intake to maximize caries prevention and minimize the occurrence of enamel fluorosis Levy 1994; Heller et al. 1999, 2000 0.05-0.07 “Useful upper limit for fluoride intake in children” Burt 1992 0.057-0.071 “Health hazard” for adults (4-5 mg/day)c McClure et al. 1945 0.057 EPA’s SMCL (2 mg/l; adult intake)i 40CFR 143.3[2001] 0.06 EPA’s reference dosej (based on protection of children from objectionable enamel fluorosis)k EPA 1989 0.083-0.13 Upper end of “safe and adequate daily dietary intake” for children 0-10 years oldd (0.5-2.5 mg/day) NRC 1989b 0.10 “Tolerable upper intake”l for ages 0-8a (0.7-2.2 mg/day) IOM 1997; ADA 2005 0.10 EPA’s SMCL (2 mg/L; child intake)m 40CFR 143.3 [2001] 0.11 EPA’s MCLG and MCL (4 mg/L; adult intake)n 40CFR 141.62(b)[2001] 0.13-0.18 “Tolerable upper intake”o for ages ≥ 14a (10 mg/day) IOM 1997; ADA 2005 0.2 EPA’s MCLG and MCL (4 mg/L; child intake)p 40CFR 141.62(b)[2001]

OCR for page 23
Fluoride in Drinking Water: A Scientific Review of EPA’S Standards Exposure, mg/kg/day Description Reference 0.25 “Tolerable upper intake”o for ages 9-13a (10 mg/day) IOM 1997; ADA 2005 aBased on intakes and average body weights listed by IOM (1997) and ADA (2005); see Table B-17 in Appendix B. bSummaries of papers published between 1979 and 1988 (IOM 1997). cBased on a 70-kg adult. dBased on intakes and median weights listed by NRC (1989b); see Table B-16 in Appendix B. eSummarized by IOM (1997). fRange, 0.045-0.056 mg/kg/day. gA minimal risk level (MRL) is an estimate of the daily human exposure to a hazardous substance that is likely to be without appreciable risk of adverse noncancer health effects over a specified duration of exposure (ATSDR 2003). hThe ATSDR (2003) states that an intermediate-duration MRL derived from a study of thyroid effects in rats would have been lower (more protective) than the chronic-duration MRL of 0.05, but the value of that MRL is not given. iBased on intake of 2 L/day by a 70-kg adult of water containing fluoride at 2 mg/L. jReference dose (RfD) is an estimate (with uncertainty spanning perhaps an order of magnitude) of a daily oral exposure to the human population (including sensitive subgroups) that is likely to be without an appreciable risk of deleterious effects during a lifetime (EPA 1989). kBased on a fluoride concentration of 1 mg/L in drinking water; the RfD for fluoride contains no uncertainty factor or modifying factor, although RfDs for other substances contain uncertainty factors to account for things such as variability within the human population (EPA 2003b). lBased on moderate enamel fluorosis (IOM 1997). mBased on intake of 1 L/day by a 20-kg child of water containing fluoride at 2 mg/L. nBased on intake of 2 L/day by a 70-kg adult of water containing fluoride at 4 mg/L. oBased on skeletal fluorosis for adults and children ≥ age 9 (IOM 1997). pBased on intake of 1 L/day by a 20-kg child of water containing fluoride at 4 mg/L. mg/kg/day as an optimum intake of fluoride but was unable to find it. He interpreted the available evidence as suggesting that 0.05-0.07 mg/kg/day (from all sources) “remains a useful upper limit for fluoride intake in children” (see also NRC 1993). Figure 2-8 shows the average intake of fluoride from all sources estimated in this report (Table 2-11), with 1 mg/L in drinking water; Figure 2-9 shows the average intake of fluoride from drinking water alone (Table 2-10), given a fluoride concentration at the MCLG/MCL (4 mg/L). For comparison purposes, an intake of 0.05-0.07 mg/kg/day is indicated on the graphs. Based on EPA’s estimates of community water consumption by consumers with an average intake (EPA 2000a), if that water is fluoridated, children

OCR for page 23
Fluoride in Drinking Water: A Scientific Review of EPA’S Standards FIGURE 2-8 Estimated average intake of fluoride from all sources, at 1 mg/L in drinking water (based on Table 2-11). Horizontal lines indicate an intake of 0.05-0.07 mg/kg/day. less than 6 months old have an intake at or above 0.05-0.07 mg/kg/day (see Appendix B, Table B-10). Children from 6 months to 1 year old have similar intakes if their water is fluoridated at 1 or 1.2 mg/L. No other age groups have that intake at ordinary fluoride concentrations; all age groups reach or exceed that intake with water at 4 mg/L. For individuals with higher-than-average intake of community water, intakes for the youngest children (<1 year) might exceed 0.05-0.07 mg/kg/day at all concentrations of water fluoridation (see Appendix B, Tables B-11, B-12, and B-13); for fluoride concentrations corresponding to the SMCL (2 mg/L) or MCL (4 mg/L), an intake of 0.05-0.07 mg/kg/day is reached or exceeded by all age groups. Note that the estimates in Appendix B include only the fluoride contribution from

OCR for page 23
Fluoride in Drinking Water: A Scientific Review of EPA’S Standards FIGURE 2-9 Estimated average intake of fluoride from drinking water alone, based on a fluoride concentration of 4 mg/L (MCLGl/MCL; based on Table 2-10). Horizontal lines indicate an intake of 0.05-0.07 mg/kg/day. community water (drinking water, plus beverages and foods prepared with community water at home or in local eating establishments); if contributions from food, tea, commercial beverages, toothpastes, and other sources are added, total intakes by individuals will increase accordingly. Estimates of total exposure (typical or average) shown in Table 2-11 indicate that all children through age 12 who take fluoride supplements (assuming low water fluoride) will reach or exceed 0.05-0.07 mg/kg/day. For children not on supplements, nonnursing infants with fluoride in tap water at ≥0.5 mg/L will exceed 0.05-0.07 mg/kg/day for typical exposures. Also, children through 5 years old (≥0.5 mg/L in tap water), children 6-12 years old (≥2 mg/L in tap water), and teenagers and adults (≥4 mg/L in tap water) will exceed 0.05-0.07 mg/kg/day with typical or average fluoride exposures in terms of water consumption and toothpaste ingestion.

OCR for page 23
Fluoride in Drinking Water: A Scientific Review of EPA’S Standards A number of researchers have pointed out both the importance of evaluating individual fluoride intake from all sources and the difficulties associated with doing so, given the variability of fluoride content in various foods and beverages and the variability of individual intakes of the specific items (Clovis and Hargreaves 1988; Nowak and Nowak 1989; Chan et al. 1990; Stannard et al. 1990, 1991; Weinberger 1991; Toumba et al. 1994; Duperon et al. 1995; Van Winkle et al. 1995; Chan and Koh 1996; Kiritsy et al. 1996; Warren et al. 1996; Heilman et al. 1997, 1999; Heller et al. 1999; Levy and Guha-Chowdhury 1999; Lalumandier and Ayers 2000). However, as shown in Figure 2-1, for typical individuals, the single most important contributor to fluoride exposures (approaching 50% or more) is fluoridated water and other beverages and foods prepared or manufactured with fluoridated water. RECOMMENDATIONS Fluoride should be included in nationwide biomonitoring surveys and nutritional studies (e.g., CDC’s National Health and Nutrition Examination Survey and affiliated studies). In particular, analysis of fluoride in blood and urine samples taken in these surveys would be valuable. National data on fluoridation (e.g., CDC 1993) should be updated on a regular basis. Probabilistic analysis should be performed for the uncertainty in estimates of individual and group exposures and for population distributions of exposure (e.g., variability with respect to long-term water consumption). This would permit estimation of the number of people exposed at various concentrations, identification of population subgroups at unusual risk for high exposures, identification or confirmation of those fluoride sources with the greatest impact on individual or population exposures, and identification or characterization of fluoride sources that are significant contributors to total exposure for certain population subgroups. To assist in estimating individual fluoride exposure from ingestion, manufacturers and producers should provide information on the fluoride content of commercial foods and beverages. To permit better characterization of current exposures from airborne fluorides, ambient concentrations of airborne hydrogen fluoride and particulates should be reported on national and regional scales, especially for areas of known air pollution or known sources of airborne fluorides. Additional information on fluoride concentrations in soils in residential and recreational areas near industrial fluoride sources also should be obtained. Additional studies on the relationship between individual fluoride exposures and measurements of fluoride in tissues (especially bone and nails) and bodily fluids (especially serum and urine) should be conducted. Such

OCR for page 23
Fluoride in Drinking Water: A Scientific Review of EPA’S Standards studies should determine both absolute intakes (mg/day) and body-weight normalized intakes (mg/kg/day). Assumptions about the influence of environmental factors, particularly temperature, on water consumption should be reevaluated in light of current lifestyle practices (e.g., greater availability of air conditioning, participation in indoor sports). Better characterization of exposure to fluoride is needed in epidemiology studies investigating potential effects. Important exposure aspects of such studies would include the following: collecting data on general dietary status and dietary factors that could influence exposure or effects, such as calcium, iodine, and aluminum intakes characterizing and grouping individuals by estimated (total) exposure, rather than by source of exposure, location of residence, fluoride concentration in drinking water, or other surrogates reporting intakes or exposures with and without normalization for body weight (e.g., mg/day and mg/kg/day) addressing uncertainties associated with exposure, including uncertainties in measurements of fluoride concentrations in bodily fluids and tissues reporting data in terms of individual correlations between intake and effect, differences in subgroups, and differences in percentages of individuals showing an effect and not just differences in group or population means. Further analysis should be done of the concentrations of fluoride and various fluoride species or complexes (especially fluorosilicates and aluminofluorides) present in tap water, using a range of water samples (e.g., of different hardness and mineral content). Research also should include characterizing any changes in speciation that occur when tap water is used for various purposes—for example, to make acidic beverages. The possibility of biological effects of SiF62−, as opposed to free fluoride ion, should be examined. The biological effects of aluminofluoride complexes should be researched further, including the conditions (exposure conditions and physiological conditions) under which the complexes can be expected to occur and to have biological effects.