a quantitative look at fluorosis, fluoride exposure, and intake in children 2005

8/10/2019 A Quantitative Look at Fluorosis, Fluoride Exposure, And Intake in Children 2005

1/7

Nearly two-thirds of the U.S. populationreceives drinking water from municipalitiesthat add fluoride to their water systems toprevent dental caries [Centers for DiseaseControl and Prevention (CDC) 2002]. The

CDC hails fluoridation of drinking water asone of the 10 great public health achieve-ments of the 20th century (CDC 1999). Thefirst Surgeon Generals report on oral healthin the United States credits fluoridation fordramatically lowering caries rates. Severalstudies have shown caries reduction of up to60% after fluoridation [U.S. Departmentof Health and Human Services (DHHS)2000a].

Although the efficacy of drinking-waterfluoridation is well accepted by the scientificcommunity and policy makers, the benefitsare not without consequence. Ingestion offluoride during the formative years of a childsenamel development can cause dental fluoro-sisa condition marked by permanent, oftenpronounced staining of adult teeth. Reports offluorosis prevalence in North American chil-dren range widely depending on public waterfluoridation status (Clark 1994; Mascarenhas2000; Riordan and Banks 1991; Tabari et al.2000). In the National Survey of DentalCaries in U.S. school children (19861987),22% of children examined had fluorosis(Brunelle 1989). In 1998, 69% of children711 years of age examined in a suburban

Boston pediatric practice were found to havefluorosis (Morgan et al. 1998). Children froma fluoridated community in North Carolinashowed a prevalence of 78% with fluorosis(Lalumandier and Rozier 1995). In nonfluori-

dated communities, fluorosis prevalencereported in a number of studies conductedduring 19902000 ranged from 3 to 45%(Clark 1994; Mascarenhas 2000; Riordan andBanks 1991; Tabari et al. 2000).

Several studies point to other sources offluoride besides fluoridated drinking water(e.g., fluoride toothpaste, fluoride supple-ments, infant formula and beverages producedwith fluoridated water, food grown in soilcontaining fluoride or irrigated with fluori-dated water, and cows milk from livestockraised on fluoride-containing water and feed,and soil) that contribute to overall fluorideintake and therefore may contribute to dentalfluorosis (Fomon et al. 2000; Jackson et al.2002; Levy 1994; Levy et al. 2001; Pendrysand Stamm 1990). In this study, we evalu-ated total fluoride intake and fluorosis risk ofinfants and children using quantitative healthrisk assessment. Although several publishedstudies in the past decade have measureddaily intake rates of fluoride from varioussources such as diet, toothpaste, and infantformula (Fomon et al. 2000; Jackson et al.2002; Levy 1994; Levy et al. 2001; Pendrysand Stamm 1990), none has systematically

considered cumulative fluoride intake fromall significant sources combined. We per-formed a comparative analysis of fluorideintake in fluoridated and nonfluoridatedcommunities by characterizing the exposuresvia all significant exposure pathways applica-ble for infants and children in two agegroups: infants less than 1 year of age andchildren 35 years of age. The analysis waslimited to formula-fed infants only.

Materials and Methods

We used the risk assessment paradigm devel-oped by the National Academy of Sciences

(1983), which is commonly used by federalenvironmental agencies in the United States toinform decisions regarding risk priorities, riskranking, and health-based environmental stan-dard development [U.S. EnvironmentalProtection Agency (EPA) 1989, 1995]. Thisrisk assessment model, in general, consists ofthe following four steps: hazard identification,doseresponse assessment, exposure assess-ment, and risk characterization. We appliedthis four-step risk assessment paradigm toquantitatively estimate exposure-pathwayspecific and cumulative daily average intake offluoride by infants and children. The signifi-cance of dental fluorosis has been controversial

at times, and there are differing perspectives bydiffering agencies and organizations chargedwith protecting the public health. Althoughthe U.S. EPA considers fluorosis a cosmeticeffect rather than an adverse health effect, theWorld Health Organization (WHO) treatsfluorosis as an adverse health effect affectingmillions of people around the world (WHO2001, 2002). In this study, we estimatedcumulative daily dose and health risk to deter-mine the exposure pathways and conditionsresulting in increased likelihood for dental flu-orosis in children, with the vision that suchinformation would be beneficial in identifyingexposure pathways of concern and in manag-ing risks for fluorosis. Therefore, application ofa quantitative risk assessment model is appro-priate for determining acceptability of risks

Address correspondence to S. Erdal, Environmentaland Occupational Health Sciences, School of PublicHealth, 2121 West Taylor St., Chicago, IL 60613USA. Telephone (312) 996-5875. Fax: (312) 413-9898. E-mail: [email protected]

The authors declare they have no competingfinancial interests.

Received 8 March 2004; accepted 14 September2004.

A Quantitative Look at Fluorosis, Fluoride Exposure, and Intake in ChildrenUsing a Health Risk Assessment Approach

Serap Erdal1 and Susan N. Buchanan2

1Division of Environmental and Occupational Health Sciences, School of Public Health, and 2Departments of Occupational Medicine andFamily Medicine, College of Medicine, University of Illinois at Chicago, Chicago, Illinois, USA

The prevalence of dental fluorosis in the United States has increased during the last 30 years. Inthis study, we used a mathematical model commonly employed by the U.S. EnvironmentalProtection Agency to estimate average daily intake of fluoride via all applicable exposure pathways

contributing to fluorosis risk for infants and children living in hypothetical fluoridated and non-fluoridated communities. We also estimated hazard quotients for each exposure pathway and haz-

ard indices for exposure conditions representative of central tendency exposure (CTE) andreasonable maximum exposure (RME) conditions. The exposure pathways considered were uptakeof fluoride via fluoridated drinking water, beverages, cows milk, foods, and fluoride supplements

for both age groups. Additionally, consumption of infant formula for infants and inadvertentswallowing of toothpaste while brushing and incidental ingestion of soil for children were alsoconsidered. The cumulative daily fluoride intake in fluoridated areas was estimated as 0.20 and

0.11 mg/kg-day for RME and CTE scenarios, respectively, for infants. On the other hand, theRME and CTE estimates for children were 0.23 and 0.06 mg/kg-day, respectively. In areas where

municipal water is not fluoridated, our RME and CTE estimates for cumulative daily averageintake were, respectively, 0.11 and 0.08 mg/kg-day for infants and 0.21 and 0.06 mg/kg-day forchildren. Our theoretical estimates are in good agreement with measurement-based estimates

reported in the literature. Although CTE estimates were within the optimum range for dentalcaries prevention, the RME estimates were above the upper tolerable intake limit. This suggeststhat some children may be at risk for fluorosis. Key words:children, exposure, fluoride, multi-

pathway, risk. Environ Health Perspect 113:111117 (2005). doi:10.1289/ehp.7077 available viahttp://dx.doi.org/ [Online 14 September 2004]

Environmental Health Perspectives

VOLUME 113|

NUMBER 1|January 2005

111

Childrens Health | Article


2/7

associated with exposure to any chemical,whether or not that chemica l has a specificadverse health effect.

Hazard identification. Several healtheffects are associated with fluoride ingestion,ranging from nausea to neurotoxic effects todeath (Mullins et al. 1998; Vogt et al. 1982).The effect of concern in our risk assessment is

the most common effect of chronic ingestionof fluoride in the form of fluoride salts: dentalfluorosis. Fluorosis occurs as permanent teethare forming and is characterized by perma-nent hypomineralization. It appears initiallyas white streaks or mottling on the toothenamel. With continued systemic exposure tofluoride, these streaks become white patches,progressing to brown stains and pitting. Theexact age at which teeth are most vulnerable issomewhat controversial, with opinions rang-ing from the prenatal stage of permanenttooth formation to 36 years of age, whenmaximal mineralization occurs. It is generallyaccepted, however, that after 68 years of age,teeth are no longer susceptible to the adverseeffects of fluoride.

Doseresponse assessment. The U.S. EPApublishes a database of toxicity values derivedfrom doseresponse relationships relating expo-sure (dose) to health effect for various chemicalsfound in the environment. This database, calledthe Integrated Risk Information System (IRIS;U.S. EPA 2003), provides the toxicity values[e.g., reference dose (RfD)] for individual non-carcinogenic chemicals. The RfD is an estimateof the daily exposure to children and adults thatis likely to be without appreciable risk of dele-terious effects during a lifetime, with uncer-

tainty spanning perhaps an order of magnitude.The RfD published by the EPA for fluorideis 0.06 mg/kg-day and is based on the noobserved adverse effects level (NOAEL) of0.06 mg/kg-day and uncertainty and modify-ing factors of unity (U.S. EPA 2003).Uncertainty factors were not deemed necessary

because NOAEL was derived from a chronicstudy focusing on the critical effect (dental flu-orosis) in a sensitive population of humans(children). The scientific basis and rationale ofthe fluoride RfD (Burt 1992; Hodge 1950)can be found in IRIS and is beyond the scopeof this article.

Exposure assessment. The populations of

interest, the pathways by which exposure mayoccur, and the magnitude, frequency, andduration of these potential exposures are iden-tified in this step. The population of interestin this analysis is infants (< 1 year of age) andchildren (35 years of age). An estimated dailyintake (EDI) is calculated for each exposurepathway using a number of exposure parame-ters using Equation 1 (U.S. EPA 1992):

[1]

where EDI is the estimated daily intake (milli-grams per kilogram per day), C is the con-centration in a specific medium (milligramsper liter or milligrams per kilogram), IR is theingestion or intake rate (milligrams per day),EF is the exposure frequency (days per year),ED is the exposure duration (years), AF is theabsorption factor (unitless), CF is the conver-sion factor (106 kg/mg), BW is the bodyweight (kilograms), and AT is the averagingtime (days).

The exposure pathways considered are asfollows: pathway A, ingestion of fluoridatedpublic drinking water; B, ingestion of softdrinks and fruit juices (beverages); C, con-sumption of infant formula; D, ingestion of

cows milk; E, consumption of foods; F, inci-dental ingestion of soil; G, ingestion of fluoridesupplement tablets; and H, incidental inges-tion of fluoride toothpaste. The exposure path-ways AE and G are included in the estimationof cumulative fluoride intake for infants. Allexposure pathways except pathway C (infant

formula) are included to estimate cumulativefluoride intake of children.

Using Equation 1, the EDI for each expo-sure pathway is calculated by identifying appro-priate values for exposure parameters (e.g.,concentration, ingestion rate, body weight,exposure frequency, exposure duration) for thetwo age groups. Two values for each exposure

parameter are used in characterizing potentialexposures: one value to represent an average orcentral tendency exposure (CTE) and anothervalue for the high-end or reasonable maximumexposure (RME), which is intended to repre-sent a plausible worst-case exposure (U.S. EPA1989). The RME estimates are often used bythe EPA when making regulatory decisions andrecommendations regarding acceptability ofhealth risk to humans.

Age-specific values used in the calculationfor EDI (Equation 1) can be found in Table 1.We consulted the U.S. EPA (2002) for theestimation of average daily fluoride intake viaall the exposure pathways except consumptionof infant formula (pathway C), ingestion offluoride supplements (pathway G), and inci-dental ingestion of toothpaste (pathway H). Amore in-depth discussion about the rationalefor exposure-pathwayspecific exposure para-meters shown in Table 1specifically, fluorideconcentrations in each exposure mediumispresented below. Exposure frequency wasassumed to be 365 days per year. Exposureduration was 1 year for infants and 2 years forchildren. Exceptions to these for EF and EDvariables in Equation 1 are also noted below.The AT is equal to ED times 365 days/year.We used average body weight of 8.4 kg for

2- to 12-month-old male and female infants,respectively, in the U.S. population, based onsurvey data from 1988 to 1994 (U.S. EPA2002), as the body weight of infants. Using thesame data source, we estimated the mean bodyweight for the 3- to 5-year-old group in asimilar fashion at 17.2 kg. The estimation of

EDI =IR EF ED AF CF

BW AT

C

,

Childrens Health | Erdal and Buchanan

112VOLUME 113

|NUMBER 1

|January 2005


Table 1. Summary of exposure parameters used in the calculation of estimated daily fluoride intake.

Exposure pathway Fluoride concentration CTE intake ratea RME intake rate

A. Drinking waterb 1 mg/L < 1 year old: 0.34 L/day < 1 year old: 0.88 L/day110 years old: 0.4 L/day (U.S. EPA 2002) 110 years old: 0.9 L/day (U.S. EPA 2002)

B. Beveragesc 0.76 mg/L (Pang et al. 1992) < 1 year old: 19 g/day < 1 year old: 23.75 g/day35 years old: 269 g/day (U.S. EPA 2002) 35 years old: 336.25 g/day (U.S. EPA 2002)

C. Infant formula 0.65 mg/kg (ATSDR 2001) 198 mL/feeding, 4.4 feedings/day 214 mL/feeding, 4.8 feedings/day

(Behrman and Vaughn 2000) (Behrman and Vaughn 2000)D. Cows milkc 0.041 mg/kg (Dabeka and McKenzie 1987) < 1 year old: 61 g/day < 1 year old: 76.25 g/day

35 years old: 335 g/day (U.S. EPA 2002) 35 years old: 418.75 g/day (U.S. EPA 2002)E. Foodb < 1 year old: 0.262 mg/kg < 1 year old: 223.6 g/day < 1 year old: 612.7 g/day

35 years old: 0.290 mg/kg 35 years old: 691.9 g/day (U.S. EPA 2002) 35 years old: 1,312.5 g/day (U.S. EPA 2002)(Dabeka and McKenzie 1995)

F. Soild 430 mg/kg (ATSDR 2001) 0.1 g/day (U.S. EPA 2002) 0.4 g/day (U.S. EPA 2002)G. Fluoride supplements 6 months to 10 years of age: 0.25 mg/day NaF 6 months to 3 years old: 0.25 mg/day NaF

36 years old: 0.5 mg/day NaF (CDC 2001) 36 years old: 0.5 mg/day NaF (CDC 2001)H. Toothpaste 1,000 mg/kg (ATSDR 2001) 35 years old: 0.26 g/brushing (Levy 1993) 35 years old: 0.77 g/brushing (Levy 1993)

1 brushing/day 3 brushings/day

aRecommended mean intake rate as a combined estimate for males and females was used in all cases in the CTE scenario. bFor drinking water and food consumption, 90th percentile ofrecommended intake rate was used in the RME scenario. cFor consumption of beverages and cows milk, 25% more consumption than the mean was assumed in the estimation of RMEdaily intake. dFor incidental ingestion of soil by children, upper percentile ingestion rate was used in the RME scenario.


3/7

EDI also requires information on absorption(or bioavailability) factor. Fluoride is readilyabsorbed from the gastrointestinal tract, withestimates of absorption ranging from 75 to100% [Agency for Toxic Substances andDisease Registry (ATSDR) 2001; Ekstrandand Ehrnebo 1980]. In toothpaste, sodiumfluoride is 100% available as fluoride ion

(Alhaique et al. 1982), and studies show alinear relationship between amount of tooth-paste ingested and serum levels of fluoride(Ekstrand and Ehrnebo 1980). Therefore, theAF in Equation 1 is assumed to be unity.

Pathway A: ingestion of drinking water.The U.S. Public Health Service sets optimaldrinking water fluoride levels based on geo-graphic temperature bands and correspondingwater consumption rates (Lalumandier andJones 1999). The recommended fluoride con-centration in temperate zones is 1 ppm, or1 mg/L. A recent study conducted by theU.S. Department of Agriculture (USDA) thatmeasured fluoride content of nationally repre-sentative municipal water samples from 24 con-solidated metropolitan statistical areas in theUnited States revealed that either water isfluoridated and contains approximately 1 mg/Lof fluoride or it is not fluoridated with un-detectable fluoride concentration (Miller-Ihliet al. 2003). The USDA study found thatapproximately 40% of the water samples werefluoridated with a mean concentration of 1.01 0.15 mg/L. We assumed that the water in non-fluoridated areas does not contain any fluoride.

The use of bottled water as the primarysource of drinking water has increased inthe United States. The American Dental

Association (ADA) recently called for labelingof fluoride concentrations on bottled waterbecause of increased use of bottled water notonly as a drinking water source but also inpreparation of infant formulas and variousfoods. Bartels et al. (2000) examined fluorideconcentrations of five commercially availablebottled water products. The results indicatedthat although there were significant differencesin fluoride concentrations among differentbrands and between different batches fromthe same brand, all products had fluorideconcentrations lower than the ADA-acceptedstandard for optimally fluoridated water(i.e., 0.71.2 mg/L dependent on the averagemaximum daily air temperature of an area).Because widespread use of bottled water is arecent phenomenon, there are limited datato ascertain how bottled water intake is affect-ing childrens teeth. Our intake/risk estimatesfor infants and children consuming non-fluoridated tap water would most likely beequivalent to intake/risk estimates of thoseconsuming bottled water as their drinkingwater source because most bottled water inthe United States is currently not fluoridated.However, this may change in the future

because of pressure from consumer groupsand federal/state regulatory agencies.

Pathway B: ingestion of soft drinks andfruit juices. The ingestion of soft drinks andcommercially prepared fruit juices has morethan doubled in the last 25 years (Levy 1994).Because these beverages are usually preparedwith fluoridated water, they can be a signifi-

cant source of fluoride. Pang et al. (1992)reported the fluoride content of sodas, juices,punches, tea, and Gatorade purchased inNorth Carolina. Fluoride levels were highlyvariable, ranging from < 0.1 to 6.7 mg/L. Weused the weighted average of these reportedconcentrations (0.76 mg/L) in calculating theEDI for this pathway.

Pathway C: consumption of infant for-mula. Infant formula processed with fluori-dated water may be a significant source offluoride in infants. In 1979, because of theconcern about fluoride intake in infants, for-mula manufacturers voluntarily agreed tolower the concentration of fluoride in theirproducts (Fomon et al. 2000; Levy 1994).However, fluoridated water used to reconsti-tute or dilute powdered or concentratedpreparations remains a concern. An averageconcentration of 0.65 mg/kg fluoride is usedin the EDI calculation. This concentrationwas derived based on a survey of fluoride con-centrations in ready-to-use formula (mean,0.23 mg/kg fluoride), concentrated liquid(mean, 0.6 mg/kg fluoride), and powderedconcentrate (1.13 mg/kg fluoride) sold inretail stores in the United States (ATSDR2001; Dabeka and McKenzie 1987). Theintake rate of infant formula was estimated

from feeding recommendations by Behrmanand Vaughn (2000).

Pathway D: ingestion of cows milk. Cowsingesting fluoridated water or feed processedwith fluoridated water produce milk contain-ing fluoride. The mean fluoride concentrationof 0.041 mg/kg (range, 0.0070.086 mg/kg)reported in a Canadian study (Dabeka andMcKenzie 1987) that surveyed fluoride con-centrations in 68 samples of milk sold in retailstores across Canadian provinces was used inthis analysis.

Human breast milk contains very low lev-els of fluoride (0.004 mg/L in nonfluoridatedand 0.01 mg/L in fluoridated areas) evenwhen intake by the mother is high (Fomonet al. 2000; Levy 1994). Moreover, the per-centage of exclusively breast-fed infants at6 months of age in the United States was only22% in 1995 (Ryan 1997). For these reasons,only formula-fed infants are included in thisanalysis. Exclusively breast-fed infants willhave a much lower average daily fluorideintake for the duration of the breast-feedingperiod than will formula-fed infants.

Pathway E: consumption of food. Dabekaand McKenzie (1995) determined fluoride

concentrations in individual food items andfood composites in various categories (milkand dairy products, meat and poultry, soups,bakery goods and cereals, vegetables, fruits andfruit juices, fats and oils, sugar and candies,beverages, and other miscellaneous items) pur-chased in Winnipeg, Canada. Food categorieswith the highest mean fluoride levels were

fish (2.118 mg/kg), soups (0.606 mg/kg), andbeverages (1.148 mg/kg). The mean fluorideconcentration in all samples, including milk,various beverages and fruit juices, and tapwater, was 0.325 mg/kg, ranging from 0.011to 4.970 mg/kg. Using these data, we esti-mated the mean fluoride concentration of0.262 and 0.29 mg/kg fluoride in foodspotentially consumed by infants and children,respectively. This estimate does not includemilk, beverages and fruit juices, and tap waterbecause these are treated separately in ouranalysis. For infants, certain food items wereexcluded from their diet (e.g., cold cuts, lunch

meat, cured meats, honey), and fluoride expo-sure due to food consumption was limited to8 months, starting at 4 months of age.

Pathway F: incidental ingestion of soil.Children inadvertently ingest soil through nor-mal hand-to-mouth behavior. Industrial sites,hazardous waste sites containing fluoride, andsoil contaminated with phosphate-containingfertilizers may have higher levels of fluoride.We used the mean fluoride concentration insoils and other surface materials in the UnitedStates (430 mg/kg; range, 1037,000 mg/kg)(ATSDR 2001) in the calculation of the EDIfor the incidental soil ingestion pathway.Because children < 1 year old are not ambula-

tory, the average daily fluoride intake for thispathway is calculated for children 35 years ofage only.

Pathway G: ingestion of fluoride supple-ments. Ingestion of fluoride supplements canbe a major exposure pathway for some chil-dren. These supplements are prescribed toinfants and children in areas that lack fluori-dated public water supplies. Although severalstudies indicate that supplements are oftenprescribed inappropriately to children influoridated areas (Lalumandier and Rozier1995; Pendrys and Katz 1989), we assumedthat only children living in nonfluoridatedareas receive supplementation. The ADA, theAmerican Academy of Pediatric Dentistry,and the American Academy of Pediatrics rec-ommend supplemental fluoride intake of0.25 and 0.5 mg/day, for children 6 monthsto 3 years of age and children 36 years of agerespectively, in areas with nonfluoridatedwater (CDC 2001). After the recommendeddosing schedule for infants, exposure was lim-ited to 6 months for infants < 1 year, startingat 6 months of age.

Pathway H: incidental ingestion of tooth-paste . Because > 90% of toothpaste sold in

Childrens Health | Fluoride exposure and risk of children


VOLUME 113|


113


4/7

North America is fluoridated, many childrenare exposed to fluoride through incidentalingestion of toothpaste. Toothpastes specifi-cally flavored for children have been linkedwith use of larger quantities of toothpaste,increasing the importance of this pathway(Levy 1994). The recommended concentra-tion for fluoride ion in the United States is

generally 1,000 mg/kg (ATSDR 2001; CDC2001). The CTE and RME ingestion rates oftoothpaste used to estimate EDI were theaverage (0.26 g toothpaste per brushing) and90th percentile (0.77 g toothpaste per brush-ing) compiled from 11 studies (CDC 2001;Levy 1993, 1994). We assumed a brushingfrequency of once daily for the CTE andthree times daily for the RME. This pathwaywas excluded from the estimation of cumula-tive fluoride intake for infants (< 1 years ofage) because several studies show that manyin this age group do not have their teethbrushed (Levy et al. 1997; Tabari et al. 2000).

The EDI representing CTE and RMEscenarios are calculated for each exposurepathway discussed above using Equation 1.Cumulative EDI of fluoride is estimated byadding EDI values for infants and childrenliving in fluoridated and nonfluoridated areasusing Equations 25. In nonfluoridated areas,fluoride concentration in drinking water wasassumed to be zero; thus, intake throughingestion of drinking water was not consid-ered. On the other hand, it was assumed thatno intake via ingestion of fluoride supple-ments would occur in fluoridated areas.

For infants living in fluoridated areas:

Cumulative EDI< 1,f= EDIA+ EDIB+ EDIC+ EDID+ EDIE. [2]

For children 35 years of age living influoridated areas:

Cumulative EDI35,f= EDIA+ EDIB+ EDID+ EDIE+ EDIF+ EDIH. [3]

For infants living in nonfluoridated areas:

Cumulative EDI< 1, nf= EDIB+ EDIC+ EDID+ EDIE+ EDIG. [4]

For children 35 years of age living innonfluoridated areas:

Cumulative EDI35, nf= EDIB+ EDID+ EDIE+ EDIF+ EDIG+ EDIH. [5]

Risk characterization. The hazard quo-tient (HQ), as an estimate of the RME andCTE health risks associated with fluorideexposure via each exposure pathway, is esti-mated by integrating exposure and toxicityinformation. The sum of the HQs, the hazardindex (HI), is then calculated by dividing

cumulative dose (EDI) by the safe dose (RfD)using Equation 6, which represents the totalfluoride intake risk:

[6]

Results

Numerical results of the CTE and RME EDIestimates for each pathway and for each agegroup are shown in Table 2. Figure 1 depictsthe RME EDI estimates for infants and chil-dren. For infants, although drinking water(52%) and infant formula (39%) are the twomost significant sources contributing to cumu-lative daily fluoride intake in fluoridated areas,infant formula (71%), fluoride supplements(13.4%), and food (12.9%) are the sources ofimportance in nonfluoridated areas. For chil-dren, toothpaste (57%), drinking water (22%),and food (9%) in fluoridated areas and tooth-paste (63%), fluoride supplements (14%), andfood (10%) in nonfluoridated areas contributesignificantly to the cumulative daily intakeunder the RME conditions.

Tables 3 and 4 show the HQ and HI val-ues estimated for infants and children for each

exposure scenario. Table 4 also documents theestimates of cumulative EDI applicable toeach exposure group living in fluoridated ornonfluoridated communities. All of the HIvalues are around unity for all CTE estimates,slightly elevated for infants living in fluori-dated areas. On the other hand, all HI esti-mates for the RME scenario are greater than

unity, indicating that cumulative daily fluor-ide intake is greater than the safe dose levelestablished by the EPA for fluoride. Therefore,there may be a segment of both populationswho are at risk of developing fluorosis due toexposure to fluoride via exposure pathwaysstudied in this analysis, under the specifiedRME conditions. Figure 2 illustrates the HIestimates for each exposure group in relationto the acceptable standard of unity for HI. It isinteresting to note that all exposure-pathwayspecific HQ values are less than or aroundunity for infants and children under CMEconditions, as shown in Table 3. However,fluoridated drinking water ingestion and con-sumption of infant formula for infants andincidental ingestion of toothpaste by childrenare associated with HQ values slightly greaterthan unity under the RME conditions, whichresult in HI estimates greater than unity.

The cumulative daily fluoride intake influoridated areas was estimated at 0.20 and0.11 mg/kg-day for RME and CTE scenarios,respectively, for infants. On the other hand,

HICumulative EDI

RfDHQ35

3535

pathway= 1= =

n

= =HICumulative EDI

RfD

HQ< 1< 1

< 1

pathway= 11

n


114VOLUME 113

|NUMBER 1

|January 2005


Table 2. EDI (mg/kg-day) estimates for CTE and RME exposure scenarios.

CTE RMEExposure pathway < 1 year old 35 years old < 1 year old 35 years old

A. Fluoridated drinking water 0.04 0.023 0.10 0.052

B. Beverages 0.0017 0.012 0.021 0.015C. Infant formula 0.067 NA 0.079 NAD. Cows milk 0.0003 0.0008 0.00037 0.001E. Food 0.0052 0.012 0.014 0.022F. Soil NA 0.0025 NA 0.01G. Fluoride supplements 0.0074 0.014 0.015 0.029H. Toothpaste NA 0.015 NA 0.13

NA, exposure pathways assumed to be not applicable.

Figure 1. RME daily intake estimates for each exposure pathway.

0.18

0.14

0.12

0.1

0.08

0.06

0.04

0.02

0

EDI(mg/k

g-day)

Drinking water Beverages Infant formula Cows milk Food Soil FI supplement Toothpaste

Exposure pathway

35 years old

< 1 year old


5/7

the RME and CTE estimates for childrenwere 0.23 and 0.06 mg/kg-day, respectively(Table 4). In areas where municipal water is notfluoridated, our RME and CTE estimates forcumulative daily average intake were, respec-tively, 0.11 and 0.08 mg/kg-day for infants and0.21 and 0.06 mg/kg-day for children. Forinfants, cumulative fluoride intake is all due

to dietary sources in fluoridated areas. In non-fluoridated areas, dietary intake constitutesabout 87% (0.1 mg/kg-day) and 90%(0.07 mg/kg-day) of the cumulative intake forinfants and about 18% (0.04 mg/kg-day) and43% (0.02 mg/kg-day) of the cumulative dailyintake for children under RME and CTE sce-narios, respectively. In fluoridated areas, dietaryintake constituted 38% (0.09 mg/kg-day) and73% (0.05 mg/kg-day) of the cumulativeintake for children for RME and CTE scenar-ios, respectively. These results demonstratethat total fluoride exposure is due mainly todietary sources for infants; however, non-dietary sources (e.g., fluoride supplements,toothpaste) gain importance for childrensexposure to fluoride. The average optimumdietary fluoride intake by children living influoridated communities is found to be close to0.05 mg/kg-day (range, 0.020.1 mg/kg-day;Institute of Medicine 1999). The Institute ofMedicine recently established a tolerable upperintake level of 0.1 mg/kg-day for infants, tod-dlers, and children 8 years of age, based onthe lowest observed adverse effect level formoderate fluorosis, using dietary fluorideintake data (Institute of Medicine 1999). All ofour dietary intake estimates fall within therange of 0.020.1 mg/kg-day, except for

infants living in fluoridated areas under theRME scenario, primarily due to ingestion ofwater. Levy et al. (2001) also found that forchildren < 12 months of age, drinking waterwas a primary source of fluoride intake.

Discussion

Several studies published in the literaturehave estimated total daily fluoride intakesfrom dietary sources and toothpaste ingestion.Pendrys and Stamm (1990) estimatedfluoride intake from diet (water and bever-ages), supplements, and toothpaste to be0.07 (range, 0.040.2) and 0.08 (range,

0.050.21) mg/kg-day for children 2 yearsof age from fluoridated and nonfluoridatedcommunities, respectively. These EDI esti-mates are similar to our estimates for 3- to5-year-old children, with the mean of0.06 mg/kg-day in both fluoridated and non-fluoridated water areas. It is interesting to notethat the high-end estimates reach approximately

0.2 mg/kg-day in both analyses. Levy et al.(2001) estimated daily fluoride intake fromwater (including beverages), toothpaste, andfluoride supplements from birth to 36 monthsof age as part of a longitudinal study in Iowa.The estimated mean intakes were 0.06 mg/kgfor the 3- to 12-month-old group and0.043 mg/kg-day for the 20- to 36-month-oldgroup. The 90th percentile values were 0.12and 0.08 mg/kg-day for infants (312 monthsof age) and 3-year-old children, respectively.On the other hand, the maximum fluorideintake estimates were significantly higher,amounting to 0.2 mg/kg-day for 3-year-oldchildren and 0.9 mg/kg-day for infants.We estimated average intake levels of 0.05 and0.06 mg/kg-day for < 1- and 3- to 5-year-oldage groups in our analysis for combined fluor-ide sources from water, beverage, fluoride sup-plement, and toothpaste, which agree with theestimates of Levy et al. (2001). Our RME EDIestimates for these four exposure pathwayswere 0.12 and 0.23 mg/kg-day for infants andchildren, respectively. Although the infant esti-mate agrees with the reported 90th percentilevalue, our RME for children is higher (close tomaximum), potentially because, although weconsider older children, Levy et al. (2001) lim-its the maximum age studied to 3 years. In

addition, Levy et al. (2001) did not includeintake of prepackaged beverages such as fruitjuice and soda, and the amount of toothpasteused and proportion ingested was estimated byparents, both of which may have led to underes-timation of fluoride intake. Jackson et al. (2002)recently estimated average daily dietary intake offluoride from food (including milk) and bever-ages using a food questionnaire and USDAintake rates for 3- to 5-year-old children livingin fluoridated and nonfluoridated towns inIndiana. The children from the fluoridatedtown had an average fluoride daily intake of0.033 mg/kg-day and a maximum intake of

0.062 mg/kg-day. On the other hand, childrenfrom nonfluoridated communities had an aver-age of 0.028 mg/kg-day and a maximum intakeof 0.058 mg/kg-day. Our EDI estimates forthese pathways (milk, food, beverages) for the3- to 5-year-old exposure group are slightlylower, with 0.04 and 0.02 mg/kg-day forRME and CTE scenarios, respectively. This

may be because our estimates for fluorideintake through milk, beverages, and food donot differentiate whether these sources comefrom fluoridated or nonfluoridated areas.

Prevention of dental caries, as establishedby numerous epidemiologic studies, has beensuch a dramatic public health achievement thatthe U.S. Public Health Service has set a goal ofincreasing the percentage of the U.S. popula-tion being served by a fluoridated supply to75%, in its Healthy People 2010 initiative(U.S. DHHS 2000b). However, the findingsof this health risk assessment study supportconcerns that a segment of the infant andchild population in the United States may beexposed to amounts of fluoride greater than theoptimum level for caries prevention. We foundthat when only dietary exposure pathwaysare considered, the EDI varies from 0.02 to0.1 mg/kg-day in nonfluoridated communities,which is within the optimum range (Institute ofMedicine 1999). However, in fluoridated com-munities, this range was 0.050.2 mg/kg-day,with drinking water and infant formula beingthe primary contributors. When nondietarysources were also considered, the cumulativeEDI values significantly increased for children,whereas there was a neg ligible differencebetween the dietary exposure and cumulative

exposure for infants. For the 3- to 5-year-oldage group, the use of fluoride supplements and,especially, inadvertent ingestion of toothpaste



VOLUME 113|


115

Table 3. The CTE and RME HQ estimates for individual exposure pathways.

CTE RMEExposure pathway < 1 year old 35 years old < 1 year old 35 years old

A. Fluoridated drinking water 0.7 0.4 1.7 0.9B. Beverages 0.03 0.2 0.04 0.2C. Infant formula 1.1 NA 1.3 NAD. Cows milk 0.005 0.01 0.006 0.02E. Food 0.09 0.2 0.2 0.4F. Soil NA 0.04 NA 0.2G. Fluoride supplements 0.1 0.2 0.2 0.5H. Toothpaste NA 0.2 NA 2.2

NA, exposure pathways assumed to be not applicable.

Table 4. HIs and cumulative EDIs (mg/kg-day) forexposure scenarios of concern.

HI Cumulative EDIExposure scenario CTE RME CTE RME

Fluoridated area< 1 year old 1.9 3.3 0.11 0.2035 years old 1.1 3.9 0.06 0.23

Nonfluoridated area< 1 year old 1.4 1.8 0.08 0.1135 years old 0.9 3.5 0.06 0.21

Figure 2. HI estimates for each exposure scenario.

4.5

4

3.5

3

2.5

2

1.5

1

0.5

0

HI

< 1 year old 35 years old < 1 year old 35 years old

RME CTE

Fluoridated

Nonfluoridated

HI = 1


6/7

containing fluoride significantly increased thetotal fluoride intake by 2- to 6-fold under theRME scenario. However, drinking water, food,fluoride supplements, and toothpaste con-tributed in similar percentages (2127%) to thecumulative EDI under the CTE scenario forchildren living in fluoridated communities.

Analysis of uncertainty is an essential com-

ponent of risk assessment. In this study, weused a single point value for each of the expo-sure parameters. However, fluoride concentra-tions in drinking water, beverages and fruitjuices, and various food items are known tovary greatly (Jackson et al. 2002). Studies meas-uring fluoride concentration in beverages donot track products to their source to verifywhether they were produced with fluoridatedwater (Heilman et al. 1999; Jack son et al.2002; Levy 1994; Pang et al. 1992). A childconsuming only beverages prepared with non-fluoridated water would have a lower fluorideintake. Children brushing more or less oftenobviously increase or decrease their risk ofswallowing toothpaste. Children using mouthrinses and gels and specially flavored tooth-paste may especially be increasing their fluor-ide intake. Because of the availability of scantdata on intake rates of bottled water, thissource was not considered. The increasedreliance on bottled water as the primary drink-ing source among the U.S. population maychange the dynamics of fluoride intake amongthe children, especially given the fact thatmany bottled water products do not containany fluoride. That is why it is paramount tocontinuously track the prevalence of dentalcaries and fluorosis at specific life stages to

determine trends and to apportion the totalintake into each source. Tea leaves containhigh levels of fluoride, and brewed tea concen-trations can range from 1 to 6 mg/L (Instituteof Medicine 1999; Pang et al. 1992). Childrengrowing up in ethnic communities with fre-quent tea consumption may have increasedhigh intake of fluoride (Cao et al. 1997; Jinet al. 2000). An epidemiologic investigationcarried out in Mexico showed that boilingwater doubled fluoride concentrations foundin nonboiled water (Grimaldo et al. 1995).Thus, food or infant formula prepared withboiled water may result in increased fluorideintake through diet. The uncertainty associatedwith concentrations of fluoride in drinkingwater, drinking water ingestion rate, consump-tion rates of beverages, cows milk, food, andfluoride supplement dosage can be classified asrelatively low because these estimates emanatefrom national-scale studies conducted by theU.S. EPA and USDA. The uncertainty for therest of the exposure parameters fall into highor medium to high categories. Fluoride con-centrations in various exposure media (e.g.,beverages, cows milk, infant formula, food,soil) and incidental toothpaste ingestion rate are

especially uncertain. Therefore, the uncertaintyin the overall intake and risk estimates can bedescribed as medium at best, most likely asmedium to high.

The HI, which considers all exposure path-ways applicable for a given exposure group,was greater than unity in all cases under theRME conditions and was within acceptable

ranges in all cases except for infants living influoridated areas under the CTE conditions.Therefore, it is likely that some infants andchildren receive fluoride levels in excess ofthose likely to be without appreciable deleteri-ous effects (U.S. EPA 2003) and are at risk forfluorosis. The findings of this study confirmthe importance of considering all potentiallyapplicable exposure pathways in estimatingcumulative daily fluoride dose for scientificallysound decision making in fluorosis risk man-agement. Although the EDI associated withthe ingestion of drinking water pathway(RME: 0.05 mg/kg-day, children; 0.1 mg/kg-day, infants; CTE: 0.02 mg/kg-day, children;0.04 mg/kg-day, infants) does not exceed theoptimum fluoride range in fluoridated areasby itself, the cumulative intake exceeds theoptimum range when other pathways are con-sidered. Therefore, one approach could beimplementation of measures to reduce fluorideintake from sources other than water in com-munities where tap water is fluoridated. Therisk management for fluorosis in these com-munities could focus on preparation of infantformula for infants and ingestion of toothpastefor children. This finding emphasizes the sig-nificance of educating parents and child-carespecialists about fluorosis risk by public health

practitioners, physicians, and dentists. Thefluorosis risk can easily be reduced by supervis-ing children while brushing and by preparinginfant formula with nonfluoridated water orpurchase of infant formula constituted with-out fluoride. A significant role in fluorosis riskmanagement is also assumed by the publichealth, medical, and dental professionals byaccurately diagnosing fluoride needs of chil-dren by inquiring about all sources that areassociated with fluoride exposure on a case-by-case basis and making informed and educateddecisions about fluoride supplement prescrip-tion unique to each child.

On the other hand, a significant findingof our analysis is that, for both age groups liv-ing in nonfluoridated areas, although underthe CTE scenario the cumulative intake iswithin the optimum range (0.06 mg/kg-dayfor children, 0.08 mg/day for infants), underthe RME scenario the cumulative intake esti-mates are higher (0.21 mg/kg-day for chil-dren, 0.11 mg/kg-day for infants), exceedingthe optimum range. This raises questionsabout the continued need for fluoridation inthe U.S. municipal water supply to protectagainst the risk of fluorosis. However, given

the uncertainties inherent in this analysis, it isnot possible to be conclusive. Further researchwith carefully designed epidemiologic studieswith enough statistical power and strong expo-sure assessment component is essential andwarranted to answer critical questions aboutthe necessity of fluoridation in the presence ofchanges in dietary behavior of children and

multiple sources of fluoride currently con-tributing total intake. Costbenefit analysis forfluoride should be a component of such stud-ies. In addition, future studies should lead tocollection of detailed exposure data for eachexposure pathway so that more robust proba-bilistic risk assessment techniques, as opposedto point estimates of intake/risk presentedhere, can be applied to obtain distribution offluoride intake/risk among children withquantitative measures of uncertainty.

REFERENCES

Alhaique F, Santucci E, Guadagnini G, de Marchi F. 1982. Is fluor-

ide in dentifrices available? Boll Chim Farm 121:573578.ATSDR. 2001. Toxicological Profile for Fluorides, Hydrogen

Fluoride, and Fluorine. Draft for Public Comment. Atlanta,

GA:Agency for Toxic Substances and Disease Registry.

Bartels D, Haney K, Khajotia SS. 2000. Fluoride concentration in

bottled water. J Okla Dent Assoc 91(1):1822.

Behrman RE, Vaughan VC, eds. 2000. Nelson Textbook of

Pediatrics. 16th ed. Philadelphia:W.B. Saunders.

Brunelle JA. 1989. The prevalence of dental fluorosis in U.S.

children, 1987 [Abstract]. J Dent Res. 68(special issue):995.

Burt BA. 1992. The changing patterns of systemic fluoride

intake. J Dent Res 71(special issue):12281237.

Cao J, Zhao Y, Liu J. 1997. Brick tea consumption as the cause of

dental fluorosis among children from Mongol, Kazak, and

Yugu populations in China [Abstract]. Food Chem Toxicol

35:827.

CDC (Centers for Disease Control and Prevention). 1999. Ten

great public health achievementsUnited States,

19001999. MMWR Morb Mortal Wkly Rep 48:241243.

CDC (Centers for Disease Control and Prevention). 2001.Recommendations for using fluoride to prevent and

control dental caries in the United States. MMWR Morb

Mortal Wkly Rep 50:142.

CDC (Centers for Disease Control and Prevention). 2002.

Populations receiving optimally fluoridated public drinking

waterUnited States, 2000. MMWR Morb Mortal Wkly

Rep 51:144147.

Clark DC. 1994. Trends in prevalence of dental fluorosis in North

America. Community Dent Oral Epidemiol 22:148152.

Dabeka RW, McKenzie AD. 1987. Lead, cadmium, and fluoride

levels in market milk and infant formulas in Canada. J Assoc

Anal Chem 70:754757.

Dabeka RW, McKenzie AD. 1995. Survey of lead, cadmium,

fluoride, nickel, and cobalt in food composites and estima-

tion of dietary intakes of these elements by Canadians in

19861988. J AOAC Int 78:897909.

Ekstrand J, Ehrnebo M. 1980. Absorption of fluoride from fluor-

ide dentifrices. Caries Res 14:96102.

Fomon SJ, Ekstrand J, Ziegler EE. 2000. Fluoride intake andprevalence of dental fluorosis: trends in fluoride intake with

special attention to infants. J Public Health Dent 60:131139.

Grimaldo M, Borja-Aburto VH, Ramrez AL, Ponce M, Rosas M,

Diaz-Barriga F. 1995. Endemic fluorosis in San Luis Potosi,

Mexico. I. Identification of risk factors associated with

human exposure to fluoride. Environ Res 68:2530.

Heilman JR, Kiritsy MC, Levy SM, Wefel JS. 1999. Assessing

fluoride levels of carbonated soft drinks. J Am Dent Assoc

130:15931599.

Hodge HC. 1950. The concentration of fluorides in drinking

water to give the point of minimum caries with maximum

safety. J Am Dent Assoc 40:436449.

Institute of Medicine. 1999. Dietary Reference Intakes for

Calcium, Magnesium, Vitamin D, and Fluoride. Washington,

DC:National Academy Press.


116VOLUME 113

|NUMBER 1

|January 2005



7/7



VOLUME 113|


117

Jackson RD, Brizendine EJ, Kelly DA, Hinesley R, Stookey GK,

Dunipace AJ. 2002. The fluoride content of foods and bev-

erages from negligibly and optimally fluoridated communi-

ties. Community Dent Oral Epidemiol 30:382391.

Jin C, Yan Z, Juanwei L, Ruodeng X, Sangbu D. 2000. Environmental

fluoride content in Tibet [Abstract]. Environ Res 83:333.

Lalumandier JA, Jones JL. 1999. Fluoride concentrations in

drinking water. J Am Water Works Assoc 91:4251.

Lalumandier JA, Rozier RG. 1995. The prevalence and risk fac-

tors of fluorosis among patients in a pediatric dental prac-

tice. Pediatr Dent 17:1925.Levy S. 1993. A review of fluoride intake from fluoride denti-

frice. J Dent Child 60:115124.

Levy SM. 1994. Review of fluoride exposures and ingestion.

Community Dent Oral Epidemiol 22:173180.

Levy SM, Kiritsy C, Slager SL, Warren JJ, Kohout FJ. 1997.

Patterns of fluoride dentifrice use among infants. Pediatr

Dent 19:5055.

Levy SM, Warren JJ, Davis CS, Kirchner HL, Kanellis MJ,

Wefel JS. 2001. Patterns of fluoride intake from birth to

36 months. J Public Health Dent 61:7077.

Mascarenhas AK. 2000. Risk factors for dental fluorosis: a

review of the recent literature. Pediatr Dent 22:269277.

Miller-Ihli NJ, Pehrsson PR, Cutrifelli RL, Holden JM. 2003. Fluoride

content of municipal water in the United States: what

percentage is fluoridated? J Food Comp Anal 16:621628.

Morgan L, Allres D, Tavares M, Bellinger D, Needleman H.

1998. Investigation of the possible associations between

fluorosis, fluoride exposure, and childhood behavior prob-

lems. Pediatr Dent 20(4):244252.

Mullins M, Warden C, Barnum D. 1998. Pediatric death and

fluoride-containing wheel cleaner [Letter]. Ann Emer Med

31:523525.

National Academy of Sciences. 1983. Risk Assessment in

Federal Government Managing the Process. Washington,

DC:National Academy Press.

Pang DTY, Phillips CL, Bawden JW. 1992. Fluoride intake from

beverage consumption in a sample of North Carolina chil-

dren. J Dent Res 71:13821388.

Pendrys DG, Katz RV. 1989. Risk of enamel fluorosis associated

with fluoride supplementation, infant formula, and fluoridedentifrice use. Am J Epidemiol 130:11991208.

Pendrys DG, Stamm JW. 1990. Relationship of total fluoride

intake to beneficial effects and enamel fluorosis. J Dent

Res 69(special issue):529538.

Riordan PJ, Banks JA. 1991. Dental fluorosis and fluoride expo-

sure in western Australia. J Dent Res 70:10221028.

Ryan AS. 1997. The resurgence of breastfeeding in the United

States. Pediatrics 99:e12.

Tabari ED, Ellwood R, Rugg-Gunn AJ, Evans DJ, Davies RM.

2000. Dental fluorosis in permanent incisor teeth in rela-

tion to w ater fluor idatio n, so cial depriv ation and toot h-

paste use in infancy. Br Dent J 189:216220.

U.S. DHHS. 2000a. Oral Health in America: A Report of the

Surgeon General. Rockville, MD:National Institute of

Dental and Craniofacial Research, National Institutes of

Health, U.S. Department of Health and Human Services.

U.S. DHHS. 2000b. Healthy People 2010Understanding and

Improving Health. 2nd ed. Washington, DC:U.S. Department

of Health and Human Services.

U.S. EPA. 1989. Risk Assessment Guidance for Superfund, Vol. 1:

Human Health Evaluation Manual (Part A). EPA/540/1-

890002. Washington, DC:U.S. Environmental Protection

Agency, Office of Emergency Response.

U.S. EPA. 1992. Guidelines for exposure assessment. Fed Reg

57:2288722938.

U.S. EPA. 1995. Guidance for Risk Characterization at the U.S.

Environmental Protection Agency. Washington, DC:U.S.

Environmental Protection Agency, Science Policy Council.

U.S. EPA. 2002. Child Specific Exposure Factors Handbook. Risk

Assessment Guidance for Superfund, Vol 1: Human HealthEvaluation Manual (Part A). EPA/540/1-890002. Washington,

DC:U.S. Environmental Protection Agency, Office of

Emergency Response.

U.S. EPA. 2003. Integrated Risk Information System. Cincinnati,

OH:U.S. Environmental Protection Agency, Environmental

Criteria and Assessment Office. Available: http:// www.

epa.gov/iris [accessed 22 October 2003].

Vogt RL, Witherell L, LaRue D, Klaucke DN. 1982. Acute fluoride

poisoning associated with an on-site fluoridator in a Vermont

elementary school. Am J Public Health 72:11681171.

WHO. 2001. Water-Related Diseases. Fluorosis: The Disease and

How It Affects People. Geneva:World Health Organization.

Available: http://www.who.int/water_sanitation_health/

diseases/fluorosis/en/ [accessed 8 December 2004].

WHO. 2002. World Water Day 2001: Oral Health. Geneva:World

Health Organization. Available: http://www.who.int/entity/

water_sanitation_health/en/oralhealth.htm [accessed

8 December 2004].

a quantitative look at fluorosis, fluoride exposure, and intake in children 2005

Documents