Fluorose dentária em crianças de São Paulo , SP , 1998-2010 Dental fluorosis in children from São Paulo , Southeastern Brazil , 1998-2010

MÉTODOS: Realizou-se análise de tendência da prevalência de fluorose dentária no período de 1998 a 2010 na cidade de São Paulo, SP. As prevalências foram calculadas para diferentes anos (1998, 2002, 2008 e 2010), a partir de dados secundários obtidos em levantamentos epidemiológicos com amostras representativas da população de 12 anos de idade. A ocorrência de fluorose foi avaliada sob luz natural utilizando o índice de Dean, preconizado pela Organização Mundial da Saúde e categorizada em normal, questionável, muito leve, leve, moderada e severa. Em 1998 foram examinadas 125 crianças; 249 em 2002; 4.085 em 2008; e 231 em 2010.

Fluoridated water and toothpastes with fl uoride are important allies in preventing dental caries, the main oral health problem in the majority of countries. 21The city of São Paulo has around 11 million inhabitants and around 99% of them have access to tap water, which has been fl uoridated since 1985.From 1988 onwards practically all toothpastes commercially available in the city contain fl uoride. 22Studies on the prevalence of dental caries show a consistent decrease in children, indicating the effectiveness of these preventative methods.However, in the context of exposure to multiple sources of fl uoride, doubts persist as to prevalence of dental fl uorosis in these children.
Products containing fl uoride may be improperly used, both in the tap water and in toothpastes.Consequently, the population's levels of dental fl uorosis, an alteration in the enamel occurring during the teeth's development due to excessive, prolonged fl uoride intake, can rise. 6 the classic pattern of occurrence (chronic endemic), the problem appears due to the intake of high concentrations of fl uoride in the drinking water, usually from wells. 6However, the modern pattern of dental fl uorosis occurrence is due to exposure to multiple sources of fl uoride, in other words, intake of products, usually water and toothpaste, containing levels of fl uoride in excess of that tolerated by the organism over long periods of time.Such levels are enough to cause milder forms of fl uorosis -in many cases unnoticeable. 12This form of fl uorosis, the population pattern of which is quite different from that exhibited in cases of chronic endemic dental fl uorosis, was denominated iatrogenic endemic fl uorosis by Narvai. 17 This is distinguished from the chronic endemic form, from an epidemiological point of view, as "mild" and "very mild" cases predominate, with low frequency of "moderate" cases and a very low frequency of "severe" cases.
Menezes et al 12 stated that the alterations classed as "very mild" and "mild" produced by fl uoride in tooth enamel are not perceived as a problem by the population.Narvai & Bighetti 15 refer to "aesthetically acceptable" forms of dental fl uorosis.With this type of dental fl uorosis (iatrogenic endemic), the public health interest is to avoid new cases occurring and maintain the inevitable cases at socially acceptable levels. 15aring in mind the existence of elevated prevalence of dental fl uorosis in some places in Brazil, 8,23,25 it was hypothesized that levels of this health problem were increasing in these locations and in the Brazilian population. 19This study aimed to evaluate this hypothesis in the state capital of São Paulo, Southeastern Brazil.

INTRODUCTION METHODS
An analysis of trends in the prevalence of dental fl uorosis between 1998 and 2010 was carried out, following the recommendations for "panel" studies.The rates of prevalence were calculated for different years (1998,  2002, 2008 and 2010), based on secondary data from epidemiological oral health surveys carried out by the State and Municipal Health Departments; and the 2002 and 2010 studies were part of a broader study known as the SBBrasil Project. 14,24Although the sampling plans were different in these population surveys, due to their different aims, the samples were considered representative of 12-year-old children in the city of São Paulo, as they met the criteria proposed by the internationally standardized World Health Organization (WHO) methodology. 24To evaluate prevalence, the dental fl uorosis index proposed by Dean 6 and recommended by the WHO was used.In all four surveys, the examinations were carried out in natural light.The examiners' training in these studies met the requisites required for epidemiological surveys of dental fluorosis. 24n 1998, 125 children were examined, 249 in 2002, 4,085 in 2008 and 231 in 2010.In order to analyze trends in prevalence, 1998 was used as the reference for comparisons, which took into consideration the respective 95% confi dence intervals at the four times.Rates of prevalence and their respective confi dence intervals of the different degrees of dental fl uorosis were also analyzed comparatively.In the "normal" and "questionable" categories, enamel appeared normal in natural light, with the surface appearing smooth, shiny and generally white or light beige.In the "very mild" category, small irregularly dispersed white and opaque patches appear on less than 25% of the tooth's surface.The "mild" category is similar to the above, but involving more than 50% of the tooth surface.In "moderate" fl uorosis, the enamel surface appears worn and there are brown stains.Fluorosis is classifi ed as severe when the surface of the enamel is severely affected and the hypoplasia is so accentuated that the anatomy of the tooth is compromised.Brown stains occur all over the surface and the tooth often appears corroded.

DISCUSSION
In the classic, 20 th century studies of dental fl uorosis, the link between dental fl uorosis and public water supplies with high levels of fl uoride was well-established. 16t was also from these studies that the correlation between low levels of fl uoride in the water and lower prevalence of dental caries was discovered. 3At almost the same time as the concept of dental fl uorosis was scientifi cally consolidated, it was discovered that appropriate levels of fl uoride in the tap water constituted an important protection factor against dental caries.This discovery radically changes public health intervention strategies for preventing and controlling dental caries.Throughout the 20 th century, the use of fl uoridated products increased, driven by the tap water and toothpastes. 1 using fl uoride constituted an effective strategy, with no damage to human health, technologies based on this scientifi c evidence grew noticeably. 21Dental fl uorosis, the only undesirable side effect of using fl uoride in these public health strategies, took on, in this context, an epidemiological pattern opposite to that described by Dean in pioneering epidemiological studies on its occurrence in communities exposed to one sole source, with elevated levels of fl uoride. 5Thus, bearing in mind the new epidemiological pattern, modern use of the expression "dental fl uorosis" requires a description of this health problem, according to the category.As at the moment it is the "very mild" and "mild" categories which predominate, it is necessary to clarify the significance of the occurrence of "moderate" and "severe" cases in each context. 17This characteristics of dental fl uorosis, in its present manifestation in the context of fl uoride added to water and toothpaste, has called into question its epidemiological relevance. 13Against this background, especially in the West, scientifi cally based objections to the use of fl uoride in public health, motivated by caution related to the possible occurrence of moderate or severe dental fl uorosis in proportions relevant in terms of population, have practically ceased.Thus, although 27.8% of the children examined in a location in Brazil had some degree of dental fl uorosis, Peres et al 20 affi rm that this was not a signifi cant factor in dissatisfaction with appearance.
The principal result of this study is that the trend of prevalence of dental fl uorosis in the city of São Paulo is stable (Figure ), with no elements to sustain the hypothesis that prevalence is increasing.This result is similar to that found in Salvador, in the fi rst decade of the 21 st century, in which no trend for increased prevalence or severity of dental fl uorosis was found. 18However, these fi ndings differ from those in Porto Alegre and Arroio do Tigre, in Rio Grande do Sul, where a prevalence was found to have increased from 7.7% to 32.6% and from 0.0% to 29.7% respectively, between 1987 and 1997. 11reover, in this study, the "very mild" and "mild" categories prevailed, a situation compatible with iatrogenic endemic fl uorosis, 17 which is characterized by the aesthetics and function of the affected teeth being uncompromised.The population characteristic of this type of fl uorosis, which affects a large number of individuals simultaneously, also differs from typical iatrogenic dental fl uorosis, as it is caused by inappropriate fl uoride intake by of one or more products containing fl uoride by a single individual. 4Although they did not qualify it in this way, Cury & Usberti 5 report a typical case of iatrogenic dental fl uorosis in Brazil, in the municipality of Piracicaba.Fejerskov et al 7 mention this type of dental fl uorosis, which they call "idiopathic", the occurrence of which is verifi ed without "apparent history of signifi cant exposure to fl uoride".The authors admit that this type is "extremely rare".
Incidentally, two aspects of the occurrence of "moderate" cases of fl uorosis, in contexts in which there was no prolonged exposure to high levels of fl uoride in wide-reaching collective vehicles (e.g. in tap water), are worth noting.The fi rst concerns the variability of the individuals with regards use of products containing fl uoride, as a consequence of incorrect prescriptions or other factors which could lead them to prolonged exposure to inappropriate levels.In a meta-analysis study covering research published in England between 1966 and 1997, it was highlighted that the use of fl uoride supplements in communities without access to fl uoridated water, during the fi rst six years of life, is associated with a signifi cantly increased risk of developing dental fl uorosis. 9The second aspect is concerned with the difficulties involved in epidemiological population based research.In such studies, it is recognized that many "cases" of dental fl uorosis (including "moderate" and "severe" cases) may have been false positives, given the diffi culties in adequately calibrating the examiners, due to the high level of subjectivity of the indices, even those used in population surveys.Add to this the fact that defects and opacities in the enamel are often recorded as dental fl uorosis when they are, effectively, no such thing.The improper inclusion of such cases of false positives constitutes a signifi cant error in many epidemiological studies of fl uorosis, conducted without proper planning.
Thus, it is necessary to consider the practical diffi culties observed in epidemiological studies of dental fl uorosis, related to the problem of diagnosis.Differentiating cases of fl uorosis from other cases in which changes occur in the enamel which are not related to fl uorosis will always be a challenge for examiners, leading to recording false positive cases.Among these alterations are: white patches due to dental caries, hypoplasia of the enamel, amelogenesis imperfecta, dentinogenesis imperfect and tetracycline stains.This means that researchers and analysts should draw their conclusions prudently and cautiously, as the results are often "contaminated" by incorrect fi gures.This is a significant limitation to the data used in this study.Another relevant limitation concerns the number of individuals examined in the surveys used in the analysis, which oscillated at the four points in time and resulted in sampling plans which differed from each other.
However, even when only the descriptive epidemiological resources and the limits mentioned are considered, the results shown in this article show the differences in prevalence and severity of dental fl uorosis in São Paulo compared with the classic epidemiological pattern.
Moreover, in order for socially acceptable levels of iatrogenic dental fl uorosis to be tolerated, ethically and aesthetically (ethical acceptance of the problem is included given the ethics of public health), 10 it is essential that monitoring activities take place, controlling the level of fl uoride which products contain, as well as epidemiological monitoring, controlling the number and types of cases in the population.Such double monitoring falls to the sanitation authorities, which are recommended, before confi rming fl uorosis, to identify the type, whether "endemic chronic" or "iatrogenic endemic" and to then adopt the measures appropriate to the type. 2 One objection which could be made against this study is concerned with the use of secondary data.As mentioned, these data were produced in epidemiological surveys which had different sampling plans, due to the different objectives for which they were carried out.The sample sizes differed markedly, varying from 125 oral examinations in 1998 to 4,085 in 2008.It could be argued that these samples were "small" in 1998 and "large" in 2008.However, these disparate samples in the four surveys produced population estimates which were statistically the same.In addition, it should be considered that 21 comparisons were made, not just two or three, and no differences were observed.If this had not been the case, the analysis would indicate statistically signifi cant differences in at least one comparison, and therefore, these different sample sizes would require caution and close attention to the conclusions drawn from data with this origin.In this case it is recognized that any limitations of the study derived from the heterogeneity of the samples, although relevant in statistical terms, may not compromise the conclusions in essential terms.Moreover, it should be borne in mind that such heterogeneity is characteristic of studies of this type, which evaluate secondary data produced from different surveys, as, in such contexts, sampling plans are rarely the same or even similar.Therefore, it needs to be recognized that comparisons refer to estimates of population parameters, both for points and for confi dence intervals.Thus, it can be stated that the samples are, more often than not, limitations in studies of this type in which analyses of trends are used based on population estimates produced from two or more cross-sectional studies, and is not unique to this analysis.For this reason, in light of the results analyzed in this study, it seems valid to recognize that the prevalence of dental fl uorosis in children in Sao Paulo can be classifi ed as stable between 1988 and 2010, overall and when considering only its "mild" and "very mild" manifestations.
prevalence of dental fl uoride in 1998, 2002, 2008 and 2010 is shown in Figure.No statistically signifi cant difference was observed in the prevalence of this health problem between the comparatively analyzed years and 1998, the reference year.The percentages according to severity of the health problem are shown in the Table and demonstrate that, on comparing all of the categories of the variable, no statistically signifi cant differences are recorded in the respective prevalence.Together, the categories "mild" + "very mild" totaled 38.4% (95%CI30.3;47.6) in 1998, 32.1% (95%CI 26.6;38.2) in 2002, 38.0% (95%CI 36.5;39.5) in 2008 and 36.4% (95%CI 30.4;42.7) in 2010.With regards the fi gure for 1998, there were no statistically signifi cant differences observed in the subsequent years.One "severe" case of fl uorosis was observed in 1998 (0.1%), but there was no record of this category of the variable in 2002, 2008 and 2010.

Table .
Number and percentage of 12-year-olds according to degree of dental fl uorosis.Municipality of São Paulo, SoutheasternBrazil, 1998, 2002, 2008and 2010.Only one case was recorded in the "severe" category, in 1998 a