DENTAL FLUOROSIS IN CHILDREN ATTENDING BASIC HEALTH UNITS

bjectives: The purpose of this study was to determine the prevalence and severity of dental fluorosis among patients attending basic health units in Londrina. Methods: Five basic health units of the urban area were randomly selected and 434 children attending these units, born between 1986 and 1989 (9 to 12 years of age), were examined. Diagnosis of dental fluorosis was performed by means of the Thylstrup and Fejerskov (TF) Index. Oral examinations were carried out by 5 previously trained dentists with the patients lying in the dental chair, under artificial light, preceded by prophylaxis, isolation with cotton rolls and air-drying of the teeth. Ten percent of the sample was reexamined and an almost perfect agreement on diagnostic criteria was obtained either on the intra-examiner or inter-examiner evaluation (K=1.00, p<0.0001). Results: The observed prevalence of dental fluorosis was 91.0%, and 87.8% of the individuals were classified as TF grade 2 or less. Conclusions: Despite of the low severity, the high prevalence observed shows the need for regular screening of dental fluorosis in Londrina; establishment of procedures to prevent the overutilization of fluoride both by the professionals and the population; sanitary surveillance of fluoride levels in water supply and also in all sources of fluoride available; besides the need for further studies to understand the factors associated to dental fluorosis among children living in Londrina. Uniterms: Dental fluorosis, epidemiology; Fluoride, adverse effects; Water fluoridation.


INTRODUCTION
The availability of fluoride has increased during the last fifty years and its employment has been almost universal.Fluoridated water and fluoride-containing products have been yielding benefits to the oral health of individuals in all age ranges and socioeconomic backgrounds.On the other hand, during the last two decades much attention has been given to the safety and risks associated to its employment.This is mainly due to the evidences demonstrating a possible increase in the prevalence of dental fluorosis in fluoridated and non-fluoridated areas all over the world 8,16,19,27,30 .Even though this increase comprises mainly mild fluorosis, there are indications of a slight increase in the moderate forms of this disturbance.Some recent reports in Brazil have demonstrated the occurrence of higher levels of dental fluorosis than expected 6, 7, 9 21, 24, 29 .
In the city of Londrina, two surveys were conducted during the 90s with a view to detect the prevalence and severity of dental fluorosis among schoolchildren from the urban area of the city 1 .The first study reported a mean prevalence of 39.5% of dental fluorosis, according to the Dean Index 11 , in schoolchildren aged 6 to 12 years old.The second study comprised 12-year-old schoolchildren and found dental fluorosis in 46.7% of the children examined.Both studies scored most of the cases as very mild or mild dental fluorosis, however moderate fluorosis was also observed in 4 subjects evaluated.
The results related to the prevalence and severity of dental fluorosis in Londrina are very similar to those reported by studies conducted in developed countries, where the disease receives more attention from the investigators and authorities, therefore showing the need to better understand the epidemiological profile of this disturbance on the different groups of the child population of Londrina.The aim of this study was to evaluate the profile of dental fluorosis on children aged 9 to 12 years old attending the dental clinics of the basic health units (BHU) of the city of Londrina, determining its prevalence and severity.

MATERIALS AND METHODS
This study was approved by the Bioethics Committee of the State University of Londrina.
The study population comprised a sample of children attending the BHUs of the urban area of the city of Londrina, born between 1986 and 1989.
The sampling procedures involved selection of the dental clinics of the BHUs and also of the children.For that purpose, it was firstly observed that the city of Londrina had a total of 21 BHUs with dental clinics located at the urban area in the year 1998 assisting the age range of interest.The total number of children enrolled who regularly attended the dental clinics was 9,600 subjects.It was also observed that the number of children did not present large variations among the clinics, what allowed the random selection of dental clinics of BHU in each area of the city (north, south, east, west and center).
The children selected for the study were those who attended the clinics during the study period.Before accomplishment of the examination, the parents or guardians of each child signed a consent form.All children in the age range of interest attending the clinics were examined until a representative number for accomplishment of the study was reached.
Determination of the sample size considered the total number of patients enrolled on the clinics (N = 9,600), an estimated prevalence of 40% 1 , and a confidence interval of 95% with a margin of error of 5%, adding up to a sample of 355 individuals.This number was increased to 400 children, therefore yielding a minimum number of 80 patients to be examined in each dental clinic.
Data collection involved accomplishment of oral examination for diagnosis of the presence or absence of dental fluorosis and classification of the positive cases according to the Thylstrup and Fejerskov (TF) Index 28 .The TF index was selected because of its higher number of classifications, therefore being more sensitive as to the variation in severity.Five dentists who worked at the randomly selected clinics attended a preparatory course for diagnosis of dental fluorosis (10 hours) and workshops for standardization and calibration of the procedures (12 hours) before performing the oral examinations on the children.
The examinations were performed after professional prophylaxis of the teeth with a brush in low-speed turbine, isolation with cotton rolls, air-drying for one minute, under artificial light and with the aid of a dental mirror number 5. Just permanent teeth presenting at least 2/3 of the clinical crown erupted were included for examination.The buccal and lingual aspects were examined and the teeth were scored as one of the ten categories of the TF Index, according to the aspect that allowed better inspection.Dental elements with extensive restorations and/or fixed orthodontic appliances were excluded from the study.Children presenting chronic systemic alterations or syndromes were excluded from the statistical analysis.The differential diagnosis between fluorosis and non-fluorosis spots was performed by means of the criteria suggested by the authors of the index adopted 28 .Data collection and statistical analysis were carried out from August 1998 up to March 1999.
Ten percent of the sample was reexamined for evaluation of intraexaminer and interexaminer agreement on diagnostic criteria.Agreement on diagnostic criteria was based on the presence or absence of dental fluorosis, by using the Kappa statistical test 17 .Significant agreement was established as a minimum criterion to perform data collection (values between 0.6 and 0.8).
The χ2 test of association and the Prevalence Rate were employed to evaluate the differences on the distribution of dental fluorosis in different subgroups.A confidence interval of 95% and a significance level of 5% (p < 0.05) were selected.

RESULTS
The Kappa coefficient values 17 achieved are shown in Table 1.
Dental fluorosis was considered present when the child had at least one pair of homologous permanent teeth classified as TF grade 1 or more.The prevalence registered according to this criterion was 91%.
Concerning the severity of dental fluorosis, it was observed that from the total 8,142 teeth examined, 37.6% were regarded as TF grade 0 or absence of fluorosis, 61.0% as TF grades 1 to 3, 0.4% as TF grades 4 to 5.There were no cases of dental fluorosis scored as TF grades 6 to 9 (Table 2).
The subjects were also classified according to the higher TF value assigned to one or more teeth on the oral cavity.As presented in Table 3, 87.8% of the children were scored as TF grade 2 or less, showing that most cases comprised a mild level of the pathology.

DISCUSSION
The prevalence of dental fluorosis observed in this study (91%) is much higher than the expected 10% for individuals exposed to fluoridated water with concentrations ranging from 0.7 to 1.2 ppm of fluoride, according to the guidelines of the World Health Organization (WHO) 23 .
Fluoride concentration in water supplies of Londrina, according to the company of water supply, varied from 0.8 to 1.1 ppm, with a mean of 0.87 ppm (± 0.08 ppm) between 1986 and 1995.This period comprises the development of the permanent dentition of the children examined and therefore the risk period for development of dental fluorosis.These values are in accordance with the Brazilian legislation 2 for the concentration of fluoride in water.Only 8% of the subjects examined did not live in Londrina during the first six years of life.However, no statistically significant difference was found in the prevalence of dental fluorosis between these subjects and those who had lived in Londrina during their early childhood.Besides the information provided by the water supply company, there were no other data available to check any variation in the fluoride levels  during the study period.
Concerning severity, scoring of the individuals according to the highest TF value observed in the oral cavity reveals that most cases were scored as the mildest forms of dental fluorosis (TF grades 1, 2 and 3).However, one case of moderate fluorosis (TF grade 4) and another of severe fluorosis (TF grade 5) were also observed.
Comparisons between the present study and other studies should be careful.One should remember that the conditions of examination of the present study were exceptional.Population-based studies are usually performed under natural light without previous air-drying.On the present study, the examinations were performed in the clinics, under artificial light and after prophylaxis and airdrying of the teeth for one minute, as suggested by the authors of the TF Index 28 .Another aspect that should be taken into account is the employment of different indexes with different procedures and diagnostic criteria (Dean Index on the other surveys an TF index on this study).Since the TF Index is a logical extension of the classification provided by the Dean Index, the results presented on this study can be transformed into the scores of the Dean Index for the purpose of comparison, by means of the model of transformation proposed by Granath, et al. 15 , and the data on the studies of Burger, et al. 5 and Mabelya, et al. 20 .The results of this transformation on the study population are presented in Table 4.
According to this classification, the prevalence of dental fluorosis among the children examined would be 41.9% and the Community Fluorosis Index (CFI) would be 0.55.Even after this transformation, the prevalence observed is still high.Compared to the children examined by Dean in Kewanee 5 , the individuals on this study presented a relative risk to develop dental fluorosis 3.4 times higher.The calculated CFI is also higher than that observed by Dean, however it is not higher than the limit of 0.6, which is considered as the threshold from which dental fluorosis would then be regarded as a matter of concern for the study population 23 .However, it should be highlighted that the questionable score was not taken into account for calculation of the CFI observed; therefore the index may be lower than expected.
In addition to the limitations of the comparisons described, the population of the present study comprised a convenient sample including individuals attending the dental services of the BHU network of Londrina.The characteristics of this population are obviously different, especially regarding the exposure to fluoride sources, considering that the current possibilities of exposure to fluoride are higher and that subjects attending dental clinics may be more exposed to fluoride, since they may be more aware of the importance of oral health control and therefore they may be led to employ caries-preventive methods containing fluoride more frequently.
The prevalence of fluorosis observed among the children attending the dental clinics of the BHUs of Londrina is also higher than the values observed by other authors who worked with population-based data instead of subjects attending clinics 16,19,25,27,30 .These studies concern the prevalence of dental fluorosis in children exposed to fluoridated water with 0.6 to 1.2 ppm of fluoride and range from 13% to 60%.However, recent studies reported levels of prevalence above 60% 8,12,13,18,22,26,32 .Three of such studies employed samples of patients attending dental services and observed high frequencies of dental fluorosis among the individuals examined.One study evaluated a population of patients attending a private pediatric dentistry clinics of North Carolina, United States, and stated that 78% of the evaluated patients aged 5 to 19 years presented signs of dental fluorosis, as measured by the Tooth Surface Index of Fluorosis (TSIF) 18 .Another study included the patients attending the pediatric dentistry clinic of the University of Iowa, United States, and reported a frequency of dental fluorosis of 72%, also through the TSIF Index 26 .Finally, Morgan et al 22 by means of the Dean Index, observed that 69% of the patients attending a pediatric dentistry clinic at the suburb of Boston, United States, demonstrated dental fluorosis.
The data available on the literature as to the prevalence of dental fluorosis in Brazilian cities with fluoridated water are recent, and many of them report levels above the expected 10% 6 .Mendonça, et al. 21 23 .Despite the impossibility to apply the results of the present study to the entire children population of the city of Londrina, they may be in agreement with other reports in the literature.Moreover, they corroborate the outcomes of two other epidemiological surveys of dental fluorosis conducted in the same city some years before, however including schoolchildren 1 .The higher prevalence reported (91%) confirms that the TF Index was more sensitive than the Dean Index for detection of the mildest manifestations of dental fluorosis, especially concerning the grade 1 of the TF Index, which may just be identified after careful cleaning and drying of the teeth, as reported by Fejerskov, et al. 14 .Considering that the cases scored as TF1 could only be detected due to the differentiated conditions for diagnosis, the detected levels of dental fluorosis are maintained stable with a prevalence around 40% on the three studies conducted in the city in the years 1995, 1996 and 1999.
However, part of the difference between the data of the present study and the other surveys conducted on the same city may also be due to the selection of a population attending dental services, as previously mentioned.
Considering the epidemiological status of dental fluorosis observed by the surveys conducted in Londrina and due to the similarity between the local phenomenon and the current findings of the national and international literature, it may be stated that levels of dental fluorosis above the expected have been affecting the children of Londrina.Since the fluoride levels reported by the water supply company of the city indicate fluoride concentrations within the limits recommended by the WHO 23 , there may be two further explanations for this increase in the occurrence of dental fluorosis.The first is related to the likely fluoride intake from other sources available to the population besides fluoridated water, such as fluoride supplements, fluoride dentifrices, topical applications of fluoride and the diet.Particularly considering fluoride dentifrices, it is necessary to remember that since 1989 these products became an important source of fluoride in Brazil, with the creation of specific legislation for the concentration of fluoride in toothpaste 2 .As a consequence of this fact, the proportion of fluoride dentifrices in Brazilian market increased from 25% in the 1988 to 90% in 1990 10 .Therefore, it must be considered that part of the children of the study was exposed to fluoridated dentifrice, besides the fluoridated water, during some of the first six years of life.The second explanation comprises the hypothesis of inadequacy of the fluoride levels in water recommended by the WHO for hot countries, as suggested by the studies conducted in Senegal 4 , Hong Kong 13 and Chile 30 .In tropical countries, the loss of liquid due to transpiration may be high; therefore, the intake of liquids due to the tropical weather may be higher 4 .Moreover, recommendations of fluoride levels in water were established on a time period on which the main fluoride sources were water and to a lesser extent the diet, different from the current availability of multiple sources of fluoride and the larger possibilities of intake.It has been suggested lately that the fluoride concentration in water supply should now range from 0.5 to 1.0 ppm of fluoride 31 .Even though this measure may not be easily accomplished in short term, since it depends on the government due to the changes required on the federal regulation currently into force, it may be achieved in middle term.In Hong Kong, for example, fluoride concentration in public water supply was adjusted many times until an ideal concentration for the local weather was found, with no damage to prevention of dental caries.

CONCLUSIONS
Despite the low severity, the high prevalence observed in this survey shows the need for regular screening of dental fluorosis in Londrina; establishment of procedures to prevent the overutilization of fluoride both by the professionals and the population; monitoring of fluoride levels in the water supply and also in all sources of fluoride available; besides the need for further studies to understand the factors associated to dental fluorosis among children living in Londrina.

TABLE 1 -
Kappa Coefficient Values before and after data collection

TABLE 2 -
Number of Teeth by TF grades in children

TABLE 3 -
Distribution of TF grades in children attending basic health units in Londrina, 1999 mentioned a prevalence

TABLE 4 -
Comparison between the TF outcomes and Dean index in children attending basic health units in Londrina, 1999

DENTAL FLUOROSIS IN CHILDREN ATTENDING BASIC HEALTH UNITS of
dental fluorosis of 69.3% in one of the six sectors of the city of Belo Horizonte.The data demonstrate that dental fluorosis is much more frequent than the 10% expected for children living in some Brazilian cities where public water supply presents fluoride concentrations within the limits recommended by WHO