Prospective Study of the Association between Fluoride Intake and Dental Fluorosis in Permanent Teeth

Original Paper DOI: 10.1159/000119520 Received: August 31, 2007 Accepted after revision: January 23, 2008 Published online: March 4, 2008 Prospective Study of the Association between Fluoride Intake and Dental Fluorosis in Permanent Teeth C.C. Martins a S.M. Paiva a Y.B. Lima-Arsati b M.L. Ramos-Jorge a J.A. Cury c a School of Dentistry, Federal University of Minas Gerais, Belo Horizonte, b São Leopoldo Mandic Dental Research Centre, Campinas, and c School of Dentistry of Piracicaba, State University of Campinas, Piracicaba, Brazil Key Words Dental fluorosis Dentifrices Fluoride Permanent dentition Abstract Objective: To evaluate the relationship between fluoride intake and dental fluorosis in permanent central incisors and first molars. Methods: Fluoride intake (mg F/kg body weight/ day) from diet, dentifrice and both combined was determined on a single occasion in 1998 among children aged 19 39 months living in two fluoridated Brazilian communities (0.6 0.8 ppm F). Six years later, when the permanent teeth of these children had erupted (central incisors and first molars), 49 children aged 7 9 years [20 girls (40.8%) and 29 boys (58.2%)] were evaluated for dental fluorosis. To test the association between fluorosis and fluoride intake, children were dichotomized into two groups, cases (children with dental fluorosis on at least two teeth, TFI 6 1) and noncases (children without dental fluorosis, TFI = 0). Results: Among the case group (n = 29), median fluoride doses from diet, dentifrice and combined were 0.031, 0.050 and 0.083 mg F/ kg/day, respectively. Among the noncase group (n = 20), median fluoride doses were 0.029, 0.049, 0.084 mg F/kg/day, respectively. There was no association between dental fluorosis in permanent teeth and fluoride intake from diet, dentifrice and combined (p 1 0.05). Conclusions: There was no difference between children with and without fluorosis in the permanent central incisors and first molars regarding fluoride intake. However, this study has limitations that must be recognized: fluoride intake was only measured once, and there were no children in the sample with severe degrees of dental fluorosis. Copyright 2008 S. Karger AG, Basel The widespread use of fluoride has led to a decrease in the prevalence of caries worldwide, but this has been accompanied by an increasing prevalence of dental fluorosis. Despite the real concern regarding dental fluorosis, there is no dose-response relationship for predicting risk. Nonetheless, the prevalence of dental fluorosis has increased with the increase in fluoride concentration in drinking water [Fejerskov et al., 1990]. One of the first reports of fluoride intake was made by McClure [1943], who estimated that the average daily diet contained 1.0 1.5 mg of fluoride, suggesting that this same diet would provide 0.05 mg F/kg body weight/day for children aged 1 12 years. This information was later interpreted as a recommendation [Farkas and Farkas, 1974]. In 1980, Ophaug et al. reported an optimum dose of 0.05 0.07 mg F/kg body weight/day. In 1992, Burt reported that the dose of 0.05 0.07 mg F/kg body weight/day began to be accepted by the scientific community as the limit dose to be considered in epidemiological studies. Fax +41 61 306 12 34 E-Mail karger@karger.ch www.karger.com 2008 S. Karger AG, Basel 0008 6568/08/0422 0125$24.50/0 Accessible online at: www.karger.com/cre Prof. Saul Martins Paiva Faculdade de Odontologia, Universidade Federal de Minas Gerais UFMG Av. Uruguai 973/402 Sion 30.310-300, Belo Horizonte, MG (Brazil) Tel. +55 31 3409 2470, Fax +55 31 3409 2472, E-Mail smpaiva@uol.com.br

A number of studies evaluating fluoride intake by young children have found that the estimated daily dose is very close to or above the limit dose [Guha-Chowdhury et al., 1990, 1996; Villa et al., 1998; Rojas-Sanchez et al., 1999; Levy et al., 2001; Paiva et al., 2003; Franco et al., 2005a; Almeida et al., 2007]. Most studies conducted in developed and developing countries have reported that fluoride intake from diet ranges from 20 to 45% of the total daily dose, while fluoride intake from dentifrices ranges from 55 to 80% of the total dose, suggesting that toothpaste has been the major contributor to fluoride intake by children [Guha-Chowdhury et al., 1990, 1996; Paiva et al., 2003; Franco et al., 2005a; Almeida et al., 2007]. In Brazil, exposure to fluoridated dentifrices increased greatly in the late 1980s. By the 1990s, 90% of dentifrices sold in Brazil were fluoridated. Recently this percentage has reached 97% [Cury et al., 2004]. Despite the considerable number of studies evaluating fluoride intake among children from diet and dentifrices, there is a lack of prospective and longitudinal studies linking data on fluoride intake and the occurrence of dental fluorosis in the permanent dentition. Only two others have been conducted thus far: one American and one carried out in seven European countries. Despite the various published manuscripts found in the literature, only the first of these studies has published data on fluorosis in the permanent dentition and its association to fluoride intake [Levy et al., 1998, 2000, 2001, 2002; Warren et al., 2002; Levy, 2003; Cochran et al., 2004a d, O Mullane et al., 2004a, b; van Loveren et al., 2004; Whelton et al., 2004; Franzman et al., 2006; Hong et al., 2006a, b]. This paper presents a prospective study on the association of fluoride intake by children and the occurrence of dental fluorosis in their permanent central incisors and first molars after a 6-year interval. Subjects and Methods Study Population The Human Research Ethics Committee of the Federal University of Minas Gerais (Brazil) approved the protocol for this research. Subjects were part of a prospective study on fluoride intake during the age of risk for dental fluorosis. The study began in 1998 with a convenience sample of 71 children from 19 to 38 months of age from the cities of Ibiá-MG and Piracicaba-SP [Paiva et al., 2003]. Both Brazilian cities are supplied with optimally fluoridated water at a concentration of 0.7 ppm F, ranging from 0.6 to 0.8 ppm F [Paiva and Cury, 2001], according to the guidelines of the Brazilian Health Ministry [1975]. Only children who regularly consumed public water took part in the study. The children from Piracicaba (n = 39) attended a public day care center and belonged to families of low socioeconomic status (SES). In order for a child to be enrolled in this day care center, the monthly income of the family should be lower than or equal to the Brazilian minimum wage (USD 190 per month). The children from Ibiá (n = 32) were recruited from a private pediatric clinic and belonged to families of high SES. In Brazil, only families of high economic status have access to private healthcare. At the time, the children were at an age of risk for the development of dental fluorosis. In 1998, fluoride intake was estimated on a single occasion from diet through the duplicate sample diet and from dentifrices through the collection of saliva expectorated after brushing (mg F/kg body weight/day) [Paiva et al., 2003]. Six years later, the children were 7 9 years old. Parents were contacted for authorization regarding dental assessments of the children. Attempts were made to locate all the families, but some had moved to other cities, some could no longer be contacted through the day care center and some telephone numbers had changed. The final sample was made up of 49 children from the original population (32 from Ibiá and 17 from Piracicaba). No subjects were excluded. All located families agreed to take part in the current investigation and the children underwent examinations for dental fluorosis in the permanent central incisors and first molars. The examination was conducted at the home of the child in accordance with the schedule of the mother and child. Central incisors and first molars were evaluated using the Thylstrup-Fejerskov Index (TFI). Fluoride Intake from Diet and Dentifrice Fluoride intake from diet was determined by the collection of duplicate portions of food and beverages consumed by children over 2 consecutive weekdays. Samples were collected by parents at home and by the researcher in the day care center. Verbal and written instructions were given to parents for the collection of duplicate amounts of all foods and beverages consumed by the children. Emphasis was placed on the importance of duplicating the diet as accurately as possible by observing what the child actually ate and drank. Parts of foods not eaten were removed, such as seeds, peels, skin and bone. All food and beverages (meals, snacks and drinks) collected during two entire days (48 h) were gathered, homogenized and analyzed for the determination of fluoride content. An average of the amount of fluoride ingested over the 2 days was calculated in order to obtain the daily fluoride intake. Fluoride from diet was extracted by the HMDS (hexamethyldisiloxane) diffusion technique [Taves, 1968]. Fluoride was analyzed using an ion-specific electrode (Orion Model 96-09, Orion Research, Cambridge, Mass., USA) and ion analyzer (Orion Model EA-940), previously calibrated with a standard solution (0.2 3.2 g F/ml). Fluoride intake from the dentifrice used was monitored maintaining the conditions at home and the day care center, with the children s toothbrushing practices observed by researcher, without interfering in the toothbrushing practice. The toothbrush was first weighed; dentifrice was squeezed onto the brush by the person usually responsible for this activity (parent, teacher or child), and brush with dentifrice was weighed again. The dentifrice used for brushing was the one that the child normally used. In both groups, only 20.0% of the children used a dentifrice specially designed for children. To carry out the analysis for fluoride content in the dentifrices, we bought the same brands used by the chil- 126 Martins /Paiva /Lima-Arsati / Ramos-Jorge /Cury

Table 1. Frequency distribution of children from the case (TFI 1) and noncase groups (TFI = 0) by age, gender, city and SES Variables TFI = 0 TFI 1 Total (n = 20) (n = 29) (n = 49) p ( 2 ) Age 7 years 3 (15.0) 4 (13.8) 7 (14.3) 0.487 8 years 17 (85.0) 23 (79.3) 40 (81.6) 9 years 0 (0.0) 2 (6.9) 2 (4.1) Gender Female 11 (55.0) 9 (31.0) 20 (40.8) 0.093 Male 9 (45.0) 20 (69.0) 29 (59.1) City Ibiá 13 (65.0) 19 (65.5) 32 (65.3) 0.970 Piracicaba 7 (35.0) 10 (34.5) 17 (34.7) SES High 13 (65.0) 19 (65.5) 32 (65.3) 0.970 Low 7 (35.0) 10 (34.5) 17 (34.7) Figures in parentheses are percentages. City and SES are the same variable. Children from Ibiá belong to high SES, and children from Piracicaba belong to low SES. dren. In both cities, the average weight of the dentifrice was 0.52 g per brushing and the average brushing frequency was 2.2 times per day. Parents/teachers were asked about how often the children brushed their teeth, and this daily toothbrushing frequency was used to calculate the daily fluoride intake from dentifrice for each child. The fluoride ingested from dentifrice at home and in the day care was combined in the Piracicaba sample. All expectorated saliva and deionized water used to rinse the mouth and toothbrush after brushing was collected. Fluoride ingestion was determined by subtracting the amount of fluoride recovered in the slurry from the amount initially placed onto the brush, thereby giving the amount of fluoride ingested on a single toothbrushing occasion. Daily fluoride intake was calculated by the fluoride ingested during toothbrushing multiplied by the daily brushing frequency reported by parents. Recovered slurry was analyzed in the laboratory to determine the fluoride content. The daily fluoride dose was determined by adding the daily fluoride intake from diet and daily fluoride intake from dentifrices. Fluoride from dentifrices was determined using an ion-specific electrode, after centrifugation, hydrolysis with 1.0 M HCl at 45 C and buffering with NaOH and TISAB II. Further information is published elsewhere [Paiva et al., 2003]. Six years following the determination of fluoride intake, the children were 7 9 years old, with central permanent incisors and first permanent molars erupted. The TFI was used for the classification of dental fluorosis in the permanent teeth [Thylstrup and Fejerskov, 1978]. The most severe degree of fluorosis on at least two teeth was considered: for example, if two teeth scored TFI = 1 and another TFI = 0, the classification for the child was TFI = 1. All 49 children were included in the study, as they all had central incisors and/or first molars erupted. Children were visited at home and were examined for dental fluorosis by a single examiner, who was blind to the fluoride intake of each child. Before the dental examination, children were asked to brush their teeth to remove any plaque or debris. Examinations were conducted at home, during daylight and teeth were dried with sterilized cotton rolls. A head lamp (PELTZ, Tikka XP, Crolles, France) and disposable mouth mirror (PRISMA, São Paulo, Brazil) were used for the dental examination. The examiner was seated in front of the child, who remained standing. A calibration program for the criteria used for the diagnosis of dental fluorosis was carried out before the study began by a single dentist (C.C.M.). Training for clinical diagnosis entailed the use of color photography to show the major clinical characteristics of each situation of interest and the situations to be considered in the differential diagnosis. Twelve children aged 7 9 years (not part of the study population) were randomly selected at schools in Ibiá and included in the calibration process. Children were examined and re-examined after a 1-week interval for the calculation of intraexaminer agreement. Cohen s -values were obtained at person level, with minimum and maximum values of 0.42 (moderate) and 0.95 (very good) [Kirkwood and Stern, 2003]. Among these 12 children, the total prevalence of dental fluorosis was 58.3% and TFI = 1 was the most prevalent degree. Informed consent terms were obtained from the parents. D a t a An a ly s i s Statistical analysis was performed employing the Statistical Package for Social Sciences (SPSS for Windows, version 12.0, SPSS Inc, Chicago, Ill., USA). In order to evaluate associations between the occurrence of fluorosis and fluoride intake, children were allocated to two groups: cases (TFI 6 1) and noncases (TFI = 0). The definition of cases was based on the presence of at least two teeth with dental fluorosis. The Shapiro-Wilk test revealed that the assumption of normality was not confirmed, so the Mann-Whitney test was used to assess differences between cases and noncases as well as between the dropout and study groups. The variables tested were the fluoride intake from diet, dentifrice and combined. This nonparametric test was chosen as fluoride intake is a continuous numerical measurement. The 2 test was used to compare the distribution of the children between the two groups, considering age, gender, city and SES. The significance level was set at = 0.05. [Kirkwood and Stern, 2003]. R e s u l t s The sample was made up of 49 children aged 7 9 years 20 girls (40.8%) and 29 boys (58.2%). The median of fluoride dose (mg F/kg body weight/day) in the group of children who were lost before reaching the age of dental fluorosis assessment (n = 22) did not differ from the final sample (n = 49; p = 0.220). The majority of children were 8 years old (n = 40, 81.6%). There was no significant difference between case and noncase groups regarding age, gender, city/ses. 59.4% of children from Ibiá and 58.9% from Piracicaba exhibited dental fluorosis; this difference was not statistically significant. Twenty-nine Association between Fluoride Intake and 127

Table 2. Median fluoride intake from diet, dentifrices and combined (mg F/kg body weight) by the children from the case (TFI 1) and noncase (TFI = 0) groups Source of fluoride TFI = 0 (n = 20) TFI 1 (n = 29) p 1 median 95% CI median 95% CI Diet 0.029 0.026 0.038 0.031 0.025 0.036 0.863 Dentifrices 0.049 0.040 0.076 0.050 0.040 0.068 0.943 Combined 0.084 0.072 0.108 0.083 0.068 0.101 0.669 1 Mann-Whitney test. children exhibited dental fluorosis on at least two teeth (cases = 59.2%), and 20 children did not exhibit any degree of fluorosis on any teeth (noncases = 40.8%; table 1 ). Among the case group, 26 children (89.7%) were classified as TFI = 1, 2 children (6.9%) were TFI = 2 and only one child (3.4%) was TFI = 4. Table 2 displays estimated fluoride intake from diet, dentifrices and combined dose for the occurrence of dental fluorosis. Noncase and case groups had similar fluoride intakes, revealing that fluoride intake was not significantly associated with the occurrence of dental fluorosis (p 1 0.05). Median values demonstrated that estimated fluoride intake from dentifrices exposed 50% of the children to a dose near or above the lower limit (0.05 mg F/kg body weight/day), with 0.049 for noncases and 0.050 for cases. Estimated fluoride intake from the combined dose was at least 0.084 (noncases) and 0.083 (cases) mg F/kg body weight/day for 50% of the children. A large portion of the children (67.3%) were exposed to a combined fluoride dose (diet + dentifrice) above the upper limit (0.07 mg F/kg body weight/day) and 83.7% of children were exposed to doses above the lower limit (0.05 mg F/kg body weight/day). Discussion Fluoride intake was determined once in 1998 among 71 children enrolled in the study, 49 of whom were located 6 years later and agreed to participate. Twenty-two families could not be located, giving an overall follow-up rate of 69.0%. No subjects were excluded and none refused to participate. A disadvantage of prospective studies is that a group of subjects is followed until the outcome of interest occurs, which can take a long time to complete and many volunteers can be lost over time [Machin and Campbell, 2005]. In the present study, however, missing data were not thought to have biased the results, as no statistically significant difference was observed between the dropout group and the final sample. There was also no statistically significant difference in sociodemographic features between the two groups enrolled in the present study ( table 1 ). There was no statistically significant difference in either age or gender between children from the case and noncase groups ( table 1 ). Most children were 8 years old (81.6%), few were 7 years old (14.3%) and only 2 were 9 years old (4.1%). Although the teeth of older children were less vulnerable to dental fluorosis at the time of the fluoride intake and dentifrice ingestion measurement, the majority of the children were in the central age (8 years old). The children from Ibiá were of high SES and those from Piracicaba were children of low SES. There was no statistically significant difference between children with or without dental fluorosis regarding SES or city ( table 1 ). We did not collect SES data at individual level, which must be recognized as a limitation of the study. However, in a study conducted in four Colombian cities all children attended day care centers, but those who attended state centers were considered of low SES, while children of high SES attended private day care centers. This Colombian study concluded that children of low SES ingested significantly more fluoride from diet and dentifrice than those from high SES [Franco et al., 2005b]. In our study, only children of low SES attended day care center. Children of high SES did not, but instead were cared for at home by parents. The present study assessed the association between children s fluoride intake during the formation of permanent central incisors and first molars and the occurrence of dental fluorosis in these teeth after eruption. A number of studies report a critical period between 2 and 3 years of age for the later development of fluorosis in the permanent anterior teeth. Thus, this period was chosen for the determination of fluoride intake in the present study (19 38 months) [Evans and Stamm, 1991; Evans and Darvell, 1995]. Studies have suggested that the first 2 years of life 128 Martins /Paiva /Lima-Arsati / Ramos-Jorge /Cury

is the most sensitive period to fluoride exposure regarding the development of dental fluorosis in the maxillary central incisors [Den Besten, 1992, 1999; Pendrys et al., 1994; Hong et al., 2006a]. Another study observed that the 1st year of life was the most critical period for the development of dental fluorosis in maxillary and mandibular central incisors [Ismail and Messer, 1996]. Fluoride intake was assessed through a duplicate sample diet and analysis of saliva expectorated after toothbrushing [Guha-Chowdhury et al., 1990, 1996]. This technique offers the advantage of analyzing the actual diet of the child without the use of food consumption tables. This method more closely represents the child s actual diet and is therefore a very accurate way of sampling the diet [Leclercq et al., 2003; Franco et al., 2005a]. Another way to estimate fluoride intake over time is to apply questionnaires to the families periodically throughout the years, as previously done in an American study [Levy et al., 2001]. The periodical administration of questionnaires has been performed in other studies with different follow-up rates: 40.3% of mothers responded to 5 questionnaires during 1 year [Levy et al., 1998], 51.0% of mothers completed questionnaires periodically and their children were clinically evaluated after about 5 years of follow-up [Warren et al., 2002], and 53.6% of mothers from the original recruited population were participating in the study after 11 years [Levy, 2003]. In the present study, we had a follow-up rate of 69.0%. The main advantage of the use of questionnaires is to get a more frequent estimate of fluoride exposure during time. Therefore, the use of questionnaires seems not to have contributed to an increased follow-up rate. The typical Brazilian meal consists of rice, beans, meat and vegetables, prepared at home. Children are rarely fed industrially processed foods, especially at ages when they are not yet enrolled at school [Buzalaf et al., 2006]. Families from Ibiá maintain the habit of preparing meals at home and the children s diet was mainly composed of these meals. Children of a high SES usually do not attend public day care centers and are cared for at home by parents, which is the case of the families from Ibiá. The children from Piracicaba attended public day care centers. These centers are supported by the Brazilian government and focus on social programs. The day care center offers three daily meals: breakfast (milk and bread with butter), lunch (rice, beans, meat and vegetables) and dinner (soup, pasta or potatoes and salad). There are no processed foods. Children eat three meals at the day care center, including dinner. In day care centers, the most common drinks offered are milk and natural fruit juices [Buzalaf et al., 2006]. The technique used to assess fluoride intake from dentifrice was the reproduction of the actual brushing practices of the children. In a European study [Cochran et al., 2004c, d], two techniques for determining fluoride intake from dentifrices were compared. First, the examiner observed the brushing event without intervening in the process. In the second method, the child kept the dentifrice tube for an entire week and fluoride intake was calculated by subtracting the weight of the tube after use from its original weight. Parents were asked to report the daily brushing frequency and dentifrice loss in order to calculate the amount of fluoride used by the child. In the first method, parents put less dentifrice on the brush than in the second method, suggesting that the presence of an observer could result in an underestimation of the dentifrice used. However, keeping the tube for a week resulted in unreported toothpaste loss and unreported dentifrice use by other family members, thereby leading to an overestimation of dentifrice use. In the present study, we decided to collect the dentifrice using the observation method, since almost all families enrolled in the study had more than one child, which could have resulted in other siblings sharing the same dentifrice tube. Furthermore, researchers have more control in determining fluoride intake when able to observe the brushing practice directly. The data that could present some information bias is the daily brushing frequency reported by parents. A study [Choi and Pak, 2005] addressing the meaning of epidemiologic biases stated that information bias can occur when asking questions on the frequency of a personal habit, which the respondent may either underestimate or overestimate. There are two prospective cohort studies on fluoride intake and dental fluorosis: The Iowa Fluoride Study in the United States and Project FLINT (Fluoride Intake from Dentifrices) in Europe. The Iowa Fluoride Study was conducted with children recruited at birth in several hospitals in the state of Iowa. Data regarding fluoride intake from dentifrices, diet and supplements were collected through questionnaires reported by parents during early childhood. At older ages, the same children were clinically examined for dental caries as well as dental fluorosis in the primary and permanent dentition. The authors found that fluorosis in permanent incisors was significantly associated with the amount of fluoride ingested from dentifrices at 24 months of age [Levy et al., 1998, 2000, 2001, 2002; Warren et al., 2002; Levy, 2003; Franzman et al., 2006; Hong et al., 2006a, b]. The FLINT Project was conducted with European children from seven countries. Fluoride intake from dentifrice was assessed Association between Fluoride Intake and 129

and data will be related to fluoride dose in the future [Cochran et al., 2004a d; O Mullane et al., 2004a, b; van Loveren et al., 2004; Whelton et al., 2004]. Fluoride intake has been assessed by a number of different methods, such as questionnaires answered by parents and data from expectorated saliva or the amount of dentifrice used by the child. Dental fluorosis has been evaluated in different ways (different indexes, examination conducted by a single examiner or by photographs), in different populations (American, European and Brazilian children), with different sample sizes and different methods of data analysis. Along with the small number of prospective researches on fluoride dose and dental fluorosis, the considerable differences between studies hinder attempts at comparison. Among the three, only the Iowa study and the present study offer data on dental fluorosis related to fluoride dose, as the European study has not yet published data on fluorosis. A large portion of the children (67.3%) were exposed to a combined fluoride dose (diet + dentifrice) above the upper limit (0.07 mg F/kg body weight/day) and 83.7% of children were exposed to doses above the lower limit (0.05 mg F/kg body weight/day). In the Iowa study, 32.0% of the children were exposed to an estimated dose below 0.04 mg F/kg body weight, 41.0% between 0.04 to 0.06 mg F/kg body weight, and 27.0% exceeded a dose of 0.06 mg F/kg body weight/day) [Hong et al., 2006b]. Children from Iowa were 0 36 months old, while children from the present study were from the same age (19 38 months). Studies evaluating only the dose obtained from the use of fluoridated dentifrice reported that 9.9% of European children [Arnadottir et al., 2004; Cochran et al., 2004d] and 19.5% of Brazilian children [Oliveira et al., 2007] are exposed to the dose limit of 0.05 0.07 mg F/kg body weight. The European study was conducted in seven countries, only one of which had a fluoridated water supply. The Brazilian study was conducted in a city with fluoridated water supply. Despite this difference, the outcomes show that many children exceed the fluoride limit dose of 0.05 0.07 mg F/kg body weight. There was no difference in median fluoride intake (diet, dentifrice and combined dose) between the case and noncase groups (p 1 0.05). Another study reported that median fluoride intake from dentifrice was significantly higher among children who exhibited dental fluorosis than children who did not exhibit fluorosis at the age of 24 months (0.017 and 0.010 mg F/kg body weight/ day) and at the age of 36 months (0.016 and 0.012 mg F/kg body weight/day). However, the median fluoride intake of children aged 16 months who exhibited dental fluorosis was not statistically different from that of children who did not exhibit fluorosis (0.002 and 0.002 mg F/kg body weight/day) [Franzman et al., 2006]. This divergence could be explained by the differences between studies, such as the sample size, method employed, data collection and the manner in which the analyses were conducted. While our fluoride intake data were obtained from laboratorial analyses, data from the Iowa study were obtained from parents reports. The sample size in Iowa was 343 children, whereas we had 49 children. To conduct a prospective cohort study using laboratory analysis for the measurement of fluoride intake is very burdensome and costly. The present study found no association between the amount of overall fluoride intake and dental fluorosis. This is contrary to the Iowa study, which reports a positive dose-response association between fluoride intake in 626 children from birth to 36 months and dental fluorosis in permanent teeth. The authors found that an average daily intake! 0.04 mg F/kg represented a low risk for fluorosis, whereas a dose of 0.04 0.06 mg F/kg represented a significantly increased risk and a daily intake 10.06 mg F/kg was associated with the highest risk for fluorosis [Hong et al., 2006b]. The studies differ in sample size, data collection, study population and data analysis. The Iowa study investigated children who were mostly from high SES and used parents responses to a questionnaire to estimate fluoride intake. We did not divide fluoride dose based on cut-off points, as most children were exposed to a fluoride dose above the limit of 0.05 0.07 mg F/kg/day. Few of our children (25%) were exposed to a fluoride dose below the lower limit (0.05 mg F/kg/day) and more than 50% of the children reached the upper limit (0.07 mg F/kg/day). Because of the small sample size the distribution made analysis based on cut-off points impossible. The fact that there was no difference between the children from the case group and noncase group regarding the amount of fluoride intake suggests that the critical limit of 0.05 0.07 mg F/kg body weight/day, should be revised and evaluated in further research. Some children were exposed to a fluoride intake above the limit dose and yet did not exhibit dental fluorosis. The association between fluoride dose and dental fluorosis is based on assumptions, as noted in the Introduction [Burt, 1992]. However, a literature review has stated that there is no critical limit for fluoride intake below which the effect on dental enamel will not occur [Aoba and Fejerskov, 2002]. Thus, the need for additional studies to more precisely define the critical limit for fluoride intake is evident. 130 Martins /Paiva /Lima-Arsati / Ramos-Jorge /Cury

Other factors may play a role in the relationship between fluoride intake and the development of fluorosis, such as the capacity of the human body to metabolize fluoride. Among the total daily fluoride intake by children aged 3 5 years old, about 35.5% is excreted through the urine and 10% through feces. Thus, the average fluoride retention is expected to be about 55% [Villa et al., 1998]. Gastric content may influence fluoride absorption as well. Ingestion of dentifrice 15 min after breakfast or after lunch reduces fluoride salivary bioavailability by 22 39.0% when compared to fasting conditions [Cury et al., 2005]. Although fluoride intake from dentifrices is higher than from diet, its absorption is decreased by gastric content, since toothbrushing is usually performed after meals. Individual variables can also influence fluoride intake and absorption. Individual fluctuations in fluoride intake over time have been reported, with greater fluctuations in the first 12 16 months, diminishing thereafter [Levy et al., 2001]. Individual biological variations (genetic and/or metabolic) may also influence the results and different children may undergo the most critical period at different times in life. Nonetheless, we do not believe that these factors had an influence over the outcome, as we had groups with similar characteristics. Additional factors that may explain the results of the present study include sample size and the fact that most cases of dental fluorosis were very mild (TFI = 1). The case and noncase groups were drawn from a prospective study and constituted a small sample that was not representative of any defined population. Perhaps if there had been more moderate and severe cases, we would have seen a greater difference between the cases and noncases regarding fluoride intake. Also, the number of participants decreased with time due to dropouts. Additional studies with a random sample and longitudinal observations evaluating the association between fluoride intake and the occurrence of dental fluorosis are advisable. Volunteers may have changed habits or acquired new ones during the follow-up period, signifying that some of the information could not be evaluated properly [Machin and Campbell, 2005]. The possibility of information bias in the parent s reporting of daily brushing frequency [Choi and Pak, 2005] may have either underestimated or overestimated the fluoride dose. In the present study, fluoride intake was evaluated on a single occasion in 1998. There was no periodical evaluation of fluoride intake during the 6-year period. Throughout this period, all children were lifelong residents of their respective cities. However, some children may have been exposed to lower or higher fluoride doses, which were not measured. At older ages, children tend to swallow less dentifrice during brushing due to an improved capacity to spit out the tooth paste, thereby decreasing the amount of fluoride ingested from dentifrice [Oliveira et al., 2007]. However, these children also drink more water and ingest more food, which may contribute to an increasing amount of fluoride from diet. Gastric content may influence fluoride absorption. Ingestion of dentifrice 15 min after breakfast or after lunch reduces fluoride salivary bioavailability when compared to fasting conditions [Cury et al., 2005]. The absorption of fluoride intake from dentifrices is decreased by gastric content, since tooth brushing is usually performed after meals. Even at the age of 19 38 months, there may be variations in fluoride intake, as the older children may swallow less dentifrice than younger children, but otherwise eat and drink more water. However, most of the children were concentrated at the age of 8 years and few were at the extremes ages. Not having documented the fluoride intake throughout the entire research period may contribute toward the limitations of the study. The present study demonstrated a lack of association between fluoride intake and the occurrence of dental fluorosis in permanent teeth. This result suggests that there is no precise limit dose above which dental fluorosis in permanent teeth will occur. The limitations of the study should be considered when interpreting the results. Further studies are required to investigate the influence of the amount of fluoride intake on the development of dental fluorosis, especially studies aimed at investigating children exposed to low fluoride doses and/or different degrees of dental fluorosis. A longitudinal approach is important for investigating the effects of increasing, decreasing or stable fluoride intake over time on tooth development and the occurrence of dental fluorosis. Acknowledgement This study was supported by Brazilian Coordination of Higher Education, Ministry of Education, Brazil. Association between Fluoride Intake and 131

References Almeida BS, Silva Cardoso VE, Buzalaf MAR: Fluoride ingestion from toothpaste and diet in 1 to 3 year-old Brazilian children. Community Dent Oral Epidemiol 2007; 35: 53 63. Aoba T, Fejerskov O: Dental fluorosis: chemistry and biology. Crit Rev Oral Biol Med 2002; 13: 155 170. Arnadottir IB, Ketley CE, van Loveren C, Seppä L, Cochran JA, Polido M, Athanossouli T, Holbrook WP, O Mullane DM: A European perspective on fluoride use in seven countries. Community Dent Oral Epidemiol 2004; 32: 69 73. Burt BA: The changing patterns of systemic fluoride intake. J Dent Res 1992; 71: 1228 1237. Buzalaf MAR, Pinto CS, Rodrigues MHC, Levy FM, Borges AS, Furlani TA, Cardoso VE: Availability of fluoride from meals given to kindergarten children in Brazil. Community Dent Oral Epidemiol 2006; 34: 87 92. Choi BCK, Pak AWP: A catalog of biases in questionnaires. Prev Chronic Dis 2005; 2: 1 13. Cochran JA, Ketley CE, Árnadóttir IB, Fernandes B, Koletsi-Kounari H, Oila AM, van Loveren C, Whelton HP, O Mullane DM: A comparison of the prevalence of fluorosis in 8-year-old children from seven European study sites using a standardized methodology. Community Dent Oral Epidemiol 2004b;32: 28 33. Cochran JA, Ketley CE, Duckworth RM, van Loveren C, Holbrook WP, Seppä L, Sanches L, Polychronopoulou A, O Mullane DM: Development of a standardized method for comparing fluoride ingested from toothpaste by 1,5 3,5-year-old children in seven European countries. 1. Field work. Community Dent Oral Epidemiol 2004c;32: 39 46. Cochran JA, Ketley CE, Duckworth RM, van Loveren C, Holbrook WP, Seppä L, Sanches L, Polychronopoulou A, O Mullane DM: Development of a standardized method for comparing fluoride ingested from toothpaste by 1,5 3,5-year-old children in seven European countries. 2. Ingestion results. Community Dent Oral Epidemiol 2004d;32: 47 53. Cochran JA, Ketley CE, Sanches L, Mamai- Homata E, Oila AM, Árnadóttir IB, van Loveren C, Whelton HP, O Mullane DM: A standardized photographic method for evaluating enamel opacities including fluorosis. Community Dent Oral Epidemiol 2004a;32: 19 27. Cury JA, Del Fiol FS, Tenuta LMA, Rosalen PL: Low fluoride dentifrice and gastrointestinal fluoride absorption alter meals. J Dent Res 2005; 84: 1133 1137. Cury JA, Tenuta LMA, Ribeiro CCC, Paes-Leme AF: The importance of fluoride dentifrices to the current dental caries prevalence in Brazil. Braz Dent J 2004; 15: 167 174. Den Besten PK: Biological mechanism of fluorosis and level and timing of systemic exposure to fluoride with respect to fluorosis. J Dent Res 1992; 71: 38 43. Den Besten PK: Mechanism and timing of fluoride effects on developing enamel. J Public Health Dent 1999; 59: 247 251. Evans RW, Darvell BW: Refining the estimate of the critical period for susceptibility to enamel fluorosis in human maxillary central incisors. J Public Health Dent 1995; 55: 238 249. Evans RW, Stamm JW: An epidemiological estimate of the critical period during which human maxillary central incisors are most susceptible to fluorosis. J Public Health Dent 1991; 51: 251 259. Farkas CS, Farkas EJ: Potential effect of food processing on the fluoride content of infant foods. Sci Total Environ 1974; 2: 399 405. Fejerskov O, Manji F, Baelum V: The nature and mechanisms of dental fluorosis in man. J Dent Res 1990; 69: 692 700. Franco AM, Martignon S, Saldarriaga A, González MC, Arbelaez MI, Ocampo A, et al: Total fluoride intake higher in subjects of low socioeconomic status. J Evid Base Dent Pract 2005b;5: 209 210. Franco AM, Martignon S, Saldarriaga A, González MC, Arbeláez MI, Ocampo A, Luna LM, Martínez-Mier EA, Villa AE: Total fluoride intake in children aged 22 35 months in four Colombian cities. Community Dent Oral Epidemiol 2005a;33: 1 8. Franzman MR, Levy SM, Warren JJ, Broffitt B: Fluoride dentifrice ingestion and fluorosis of the permanent incisors. J Am Dent Assoc 2006; 137: 645 652. Guha-Chowdhury N, Brown RH, Shepherd MG: Fluoride intake of infants in New Zealand. J Dent Res 1990; 69: 1828 1833. Guha-Chowdhury N, Drummond BK, Smillie AC: Total fluoride intake in children aged 3 to 4 years: a longitudinal study. J Dent Res 1996; 75: 1451 1457. Health Ministry of Brazil. Decree 635 December 26, 1975 (in Portuguese). Hong L, Levy SM, Broffitt B, Warren JJ, Kanellis MJ, Wefel JS, Dawson DV: Timing of fluoride intake in relation to development of fluorosis on maxillary central incisors. Community Dent Oral Epidemiol 2006a;34: 299 309. Hong L, Levy SM, Warren JJ, Broffitt B, Cavanaugh J: Fluoride intake levels in relation to fluorosis development in permanent maxillary central incisors and first molars. Caries Res 2006b;40: 494 500. Ismail AI, Messer JG: The risk of fluorosis in students exposed to a higher than optimal concentration of fluoride in well water. J Public Health Dent 1996; 56: 22 27. Kirkwood BR, Stern J: Essentials of Medical Statistics. London, Blackwell, 2003, p 512. Leclercq C, Arcella D, Armentia A, Boon PE, Kruizinga AG, Gilsenan MB, Thompson RL: Development of databases for use in validation studies of probabilistic models of dietary exposure to chemicals and nutrients. Food Addit Contam 2003; 20:S27 S35. Levy SM: An update on fluorides and fluorosis. J Can Dent Assoc 2003; 69: 286 291. Levy SM, Hillis SL, Warren JJ, Broffitt BA, Mahbubul Islam AKM, Wefel JS, Kanellis MJ: Primary tooth fluorosis and fluoride intake during the first year of life. Community Dent Oral Epidemiol 2002; 30: 286 295. Levy SM, Kiritsy MC, Slager SL, Warren JJ: Pattern of dietary fluoride supplement use during infancy. J Public Health Dent 1998; 58: 228 233. Levy SM, Mcgrady JA, Bhuridej P, Warren JJ, Heilman JR, Wefel JS: Factors affecting dentifrice use and ingestion among a sample of U.S. preschoolers. Pediatr Dent 2000; 22: 389 394. Levy SM, Warren JJ, Davis CS, Kirchner HL, Kanellis MJ, Wefel JS: Patterns of fluoride intake from birth to 36 months. J Public Health Dent 2001; 61: 70 77. Machin D, Campbell MJ: Surveys, cohort and case-control studies; in Machin D, Campbell MJ: Design of Studies for Medical Research. New York, Wiley, 2005, pp 109 138. McClure FJ: Ingestion of fluoride and dental caries: quantitative relations based on food and water requirements of children 1 to 2 years old. Am J Dis Child 1943; 66: 362 369. Oliveira MJL, Paiva SM, Martins LHPM, Ramos-Jorge ML, Lima YBO, Cury JA: Fluoride intake by children at risk for the development of dental fluorosis: comparison of regular dentifrices and flavoured dentifrices for children. Caries Res 2007; 41: 460 466. O Mullane DM, Cochran JA, Whelton HP: Fluoride ingestion from toothpaste: background to European Union-funded multicentre project. Community Dent Oral Epidemiol 2004a;32: 5 8. O Mullane DM, Ketley CE, Cochran JA, Whelton HP, Holbrook WP, van Loveren C, Fernandes B, Seppä L, Athanassouli T: Fluoride ingestion from toothpaste: conclusions of European Union-funded multicentre project. Community Dent Oral Epidemiol 2004b;32: 74 76. Ophaug RH, Singer L, Harland BF: Dietary fluoride intake of 6-month and 2-year-old children in four dietary regions of the United States. J Dent Res 1980; 59: 777 781. Paiva SM, Cury JA: Longitudinal evaluation of a fluoridation system of a public water supply in Brazil. J Dent Res 2001; 80: 1028. Paiva SM, Lima YBO, Cury JA: Fluoride intake by Brazilian children from two communities with fluoridated water. Community Dent Oral Epidemiol 2003; 31: 184 191. 132 Martins /Paiva /Lima-Arsati / Ramos-Jorge /Cury

Pendrys DG, Katz RV, Morse DE: Risk factors for enamel fluorosis in a fluoridated population. Am J Epidemiol 1994; 140: 461 471. Rojas-Sanches F, Kelly SA, Drake KM, Eckert GJ, Stookey GK, Dunipace AJ: Fluoride intake from foods, beverages and dentifrice by young children in communities with negligibly and optimally fluoridated water: a pilot study. Community Dent Oral Epidemiol 1999; 27: 288 297. Taves DR: Separation of fluoride by rapid diffusion using hexamethyldisiloxane. Talanta 1968; 15: 969 974. Thylstrup A, Fejerskov O: Clinical appearance of dental fluorosis in permanent teeth in relation to histologic changes. Community Dent Oral Epidemiol 1978; 6: 315 328. van Loveren C, Ketley CE, Cochran JA, Duckworth RM, O Mullane DM: Fluoride ingestion from toothpaste: fluoride recovered from toothbrush, the expectorate and the after-brush rinses. Community Dent Oral Epidemiol 2004; 32: 54 61. Villa AE, Guerreros S, Icaza G, Villalobos J, Anabalón M: Dental fluorosis in Chilean children: evaluation of risk factors. Community Dent Oral Epidemiol 1998; 26: 310 315. Warren JJ, Levy SM, Kanellis MJ: Dental caries in the primary dentition: assessing prevalence of cavitated and noncavitated lesions. J Public Health Dent 2002; 62: 109 114. Whelton HP, Ketley CE, Mcsweeney F, O Mullane DM: A review of fluorosis in the European Union: prevalence, risk factors and aesthetic issues. Community Dent Oral Epidemiol 2004; 32: 9 18. Association between Fluoride Intake and 133