Inequalities in untreated dental caries prevalence in preschool children in Brazil

OBJECTIVE: To evaluate the infl uence of social inequalities of individual and contextual nature on untreated dental caries in Brazilian children. METHODS: The data on the prevalence of dental caries were obtained from the Brazilian Oral Health Survey (SBBrasil 2010) Project, an epidemiological survey of oral health with a representative sample for the country and each of the geographical micro-regions. Children aged fi ve (n = 7,217) in 177 municipalities were examined and their parents/guardians completed a questionnaire. Contextual characteristics referring to the municipalities in 2010 (mean income, fl uorodized water and proportion of residences with water supply) were supplied by the Fundação Instituto Brasileiro de Geografi a e Estatística (Brazilian Institute of Geography and Statistics). Multilevel Poisson regression analysis models were used to assess associations. RESULTS: The prevalence of non-treated dental caries was 48.2%; more than half of the sample had at least one deciduous tooth affected by dental caries. The index of dental caries in deciduous teeth was 2.41, with higher fi gures in the North and Northeast. Black and brown children and those from lower income families had a higher prevalence of untreated dental caries. With regards context, the mean income in the municipality and the addition of fl uoride to the water supply were inversely associated with the prevalence of the outcome. CONCLUSIONS: Inequalities in the prevalence of untreated dental caries remain, affecting deciduous teeth of children in Brazil. Planning public policies to promote oral health should consider the effect of contextual factors as a determinant of individual risk. DESCRIPTORS: Child. Dental Caries, epidemiology. Socioeconomic Factors. Health Inequalities. Dental Health Surveys. Oral Health. Original Articles DOI:10.1590/S0034-8910.2013047004352 2 Inequality and dental caries in pre-school children Ardenghi TM et al Applying measures which are effective in preventing and controlling disease requires the factors which infl uence the population’s state of health to be identifi ed. Social inequalities have been highlighted as an important determinant in the health-disease process and have been recognized at the level of the individual, as well as at the level of social context.14 The link between socioeconomic conditions and the prevalence of dental caries has been highlighted in studies which interpret this association as the source of inequalities in oral health.1,7,19 This affi rmation has also been made for children of pre-school age.21 In this age group, the early development of dental caries is considered a public health problem, due to its high prevalence and because it affects the individual’s quality of life.11 The relationship between individual determinants and different health conditions, dental caries among them, is infl uenced by the context in which the individuals dwell.10 Previous studies have discussed the occurrence of these health problems as being related to a complex causal network which includes determinants related to individual characteristics and the social and geographical context in hierarchically organized levels.1,4,9 The action of these factors on the risk of disease does not occur in isolation; on the contrary, multiple interactions should be refl ected in conceptual models which organize the distal and proximal levels in order to explain the disease’s distribution.10 Analyzing epidemiological data without considering variance in the hierarchical levels on which the population is organized may be source of bias and lead to erroneous conclusions.1,24 To deal with this diffi culty, recent studies have used multilevel analysis in order to integrate individual characteristics and geographical variables into one explanatory model.1,4,9,13 In Brazil, epidemiological surveys assessing the oral health of the population have been carried out on a national scale since the 1980s. Analysis of the results of these surveys has highlighted the infl uence of contextual factors, incorporating socioeconomic and environmental indicators such as the Human Development Index, measures of access to health care services and the availability of a fl uoridated water supply. Moreover, these surveys have shown a decrease in dental caries indices in 12-year-old children.17 However, few studies have evaluated the distribution of early childhood caries and its associated individual and contextual factors.1,3,13 This study aimed to assess the influence of social inequalities at the individual and contextual level on untreated dental caries in Brazilian children. Identifying INTRODUCTION the current epidemiological situation of dental caries in fi ve-year-old children may aid in defi ning public policies for oral health aiming at decreasing social differences.

Geography and Statistics).Multilevel Poisson regression analysis models were used to assess associations.

RESULTS:
The prevalence of non-treated dental caries was 48.2%; more than half of the sample had at least one deciduous tooth affected by dental caries.The index of dental caries in deciduous teeth was 2.41, with higher fi gures in the North and Northeast.Black and brown children and those from lower income families had a higher prevalence of untreated dental caries.With regards context, the mean income in the municipality and the addition of fl uoride to the water supply were inversely associated with the prevalence of the outcome.

CONCLUSIONS:
Inequalities in the prevalence of untreated dental caries remain, affecting deciduous teeth of children in Brazil.Planning public policies to promote oral health should consider the effect of contextual factors as a determinant of individual risk.

DESCRIPTORS: Child. Dental Caries, epidemiology. Socioeconomic Factors. Health Inequalities. Dental Health Surveys. Oral Health.
Applying measures which are effective in preventing and controlling disease requires the factors which infl uence the population's state of health to be identifi ed.Social inequalities have been highlighted as an important determinant in the health-disease process and have been recognized at the level of the individual, as well as at the level of social context. 14e link between socioeconomic conditions and the prevalence of dental caries has been highlighted in studies which interpret this association as the source of inequalities in oral health. 1,7,19This affi rmation has also been made for children of pre-school age. 21In this age group, the early development of dental caries is considered a public health problem, due to its high prevalence and because it affects the individual's quality of life. 11e relationship between individual determinants and different health conditions, dental caries among them, is infl uenced by the context in which the individuals dwell. 10Previous studies have discussed the occurrence of these health problems as being related to a complex causal network which includes determinants related to individual characteristics and the social and geographical context in hierarchically organized levels. 1,4,9The action of these factors on the risk of disease does not occur in isolation; on the contrary, multiple interactions should be refl ected in conceptual models which organize the distal and proximal levels in order to explain the disease's distribution. 10Analyzing epidemiological data without considering variance in the hierarchical levels on which the population is organized may be source of bias and lead to erroneous conclusions. 1,24To deal with this diffi culty, recent studies have used multilevel analysis in order to integrate individual characteristics and geographical variables into one explanatory model. 1,4,9,13 Brazil, epidemiological surveys assessing the oral health of the population have been carried out on a national scale since the 1980s.Analysis of the results of these surveys has highlighted the infl uence of contextual factors, incorporating socioeconomic and environmental indicators such as the Human Development Index, measures of access to health care services and the availability of a fl uoridated water supply.Moreover, these surveys have shown a decrease in dental caries indices in 12-year-old children. 17However, few studies have evaluated the distribution of early childhood caries and its associated individual and contextual factors. 1,3,13

INTRODUCTION
the current epidemiological situation of dental caries in fi ve-year-old children may aid in defi ning public policies for oral health aiming at decreasing social differences.

METHODS
This cross-sectional study used information on the prevalence of dental caries provided by the epidemiological Pesquisa Nacional de Saúde Bucal (Projeto SBBrasil 2010 -Brazilian Oral Health Survey).The methodological procedures used in this survey, as well as the technical report are available online.a The methodology used by the SBBrasil 2010 followed international standards set by the World Health Organization (WHO) 26 and aimed to evaluate deciduous teeth of a sample of fi ve-year-olds, representative at the national level, each state capital and towns in the interior of each macro-region.
The SBBrasil 2010 carried out oral exams on 7,217 five-year-old children in 177 municipalities and applied a questionnaire to their parents or guardians.The sample was obtained using a complex cluster sampling design, and the respective sampling weights were recorded in the database.The primary sampling unit was the cluster, constituted of census tracts in the state capitals and hinterland towns in each macro region, while the children examined were the secondary sampling unit.This database enabled the prevalence and severity of the main oral health problems to be evaluated, as well as the distribution of the sociodemographic factors in question.
The indicators of dental caries prevalence were assessed using the dmft index (decayed, missing or fi lled teeth -deciduous teeth).The prevalence of untreated dental caries was defi ned by the appearance of at least one deciduous tooth with this condition, that is, with the component 'd' of the dmft index equal to or greater than one.The examiners were trained and instructed in using the criteria proposed by the WHO (1997) 26 for diagnosing dental caries.The principal outcome variable in this study was the prevalence of untreated dental caries, into which was integrated information on the prevalence of the disease and the lack of dental treatment. 3dividual-level sociodemographic characteristics were gathered using the SBBrasil 2010 questionnaire: sex, skin color and household income.The contextual characteristics of the municipalities were obtained by consulting the results of the 2010 IBGE census.This included indicators of household income (median for the municipality) and the percentage of residences connected to the water supply (categorized by quartiles of distribution).The National Basic Sanitation Survey, carried out by the IBGE 12 showed, as of 2008, which towns did and did not add fl uoride to the public water supply.
The data were analysed using the Stata 12 program.Firstly, descriptive analysis was carried out for the dmft index, prevalence of dental caries (dmft ≥ 1) and the prevalence of untreated dental caries (d ≥ 1) for each state capital and for the towns in the interior of each macro-region.Poisson multilevel regression analysis was used to evaluate the link between indicators of dental caries and the individual and contextual factors.Both the descriptive estimates and the analyses were carried out considering the complex cluster sample design and the respective sampling weight.To assess association between variables, prevalence ratios and confi dence intervals (95%) were used.
Mixed effect multilevel analysis was used, 24 estimating the fi xed effect of the measures of association between the outcome and the fi rst level factors (children), and the random effect between the outcome and the second level factors (clusters of the sampling design).In the bivariate analysis, the two outcome were compared: the prevalence of dental caries and the prevalence of untreated dental caries.In the multiple regression analysis, only one outcome was selected: the prevalence of untreated dental caries.
The multiple analysis was carried out in three stages.First, the empty mode ("model 1") was estimated, without factors, only sharing the variance in the two levels of analysis."Model 2" included only factors at the individual level."Model 3" included all of the factors adjusted according to the individual and contextual variables.The models' goodness of fi t was evaluated using the statistic -2 log likelihood; signifi cant changes in the goodness of fi t were analyzed using the likelihood ratio test. 24e SBBrasil 2010 Project followed the standards set by the Declaration of Helsinki and was approved by the Conselho Nacional de Ética em Pesquisa, record no. 15,498, 7 th January 2010.

RESULTS
A total of 7,217 fi ve-year-old children participated in the study; the no-response rate (sample loss) was 1.8%.Dental caries indices were markedly unequal in their geographic distribution.The prevalence of untreated caries, for example, varied between 27.0% in Rio de Janeiro to 68.9% in towns in the interior of the Northeast.Overall, the poorer regions of the country (the North and Northeast) had worse indices of dental caries in deciduous teeth.Compared with the Southeast, the North had a dmft index around 63% higher.In each macro-region, the towns in the interior had worse indices of dental caries than the respective state capitals.With regards the prevalence of untreated dental caries, the lowest percentages were obtained in the state capitals in the South and Southeast (Figure 1, Table 1).
The same factors which are associated with the prevalence of dental caries (dmft ≥ 1) are also associated with the prevalence of untreated dental caries (d ≥ 1).At the individual level, skin color and household income were the socio-demographic characteristics signifi cantly associated with the two measures.At the contextual level, the indices of dental caries were associated with: fl uoride added to the water supply, percentage of households connected to the water supply and median household income in the municipality (Table 2).
In the multilevel analysis, adjusted for sex, it was observed that children with brown and yellow skin had, respectively, 9% (95%CI 1.0%;17.0%)and 26% (95%CI 1.0%;58.0%)higher prevalence of dental caries than children with white skin.It was also observed

Southeast South Central-West
Northeast that, in poorer children (household income below R$ 500.00), the prevalence of untreated dental caries was 2.45 (95%CI 2.07;2.93)times higher than that of children from better off families (household income over R$ 2,500.00)(Table 3).
When contextual variables are fi tted in the multilevel model, the prevalence of the outcome was signifi cantly lower in municipalities which add fluoride to the public water supply (prevalence ratio, PR 0.88; 95%CI 0.79;0.98)and in those municipalities with higher  3).

DISCUSSION
The report of high prevalence of untreated dental caries (48.2%) in fi ve-year-old Brazilian children is consistent with research 18 showing that, in 2008, only one third of children (33.2%) of children under six had seen a dentist at least once in their lives.On the other hand, although high, the prevalence of untreated dental caries was lower than the estimate for 2003, which was 54.0%. 3 With regards the dmft index, the national estimate for 2010 (2.41) corresponds to a 14% decrease, when compared with that observed in 2003 (2.80). 3 addition to documenting actual levels of dental treatment needs in deciduous teeth, this study assessed the inequality in its distribution between Brazilian regions and between children with different socio-demographic characteristics.The epidemiological situation for dental caries was worse for children living in poorer regions of the country; the association between socio-economic conditions and the disease remained even after adjusted for individual characteristics.These results corroborate previous assessments 1,3,13 and show that socio-economic inequalities relative to the context are still important factors in infl uencing oral health in pre-school aged children.Inequalities in oral health care have been demonstrated even against the background of decreasing prevalence of dental caries over the last few decades 17 and have been described as one of the great public health challenges to be faced. 23In this sense, carrying out studies which include individual and contextual determinants to explain the distribution of dental caries are suggested as a propitious strategy in guiding initiatives aimed at reducing inequalities in health care. 23This study identifi ed racial and socioeconomic differences as individual determinants in untreated dental caries.Racial differences in dental caries indices have been reported in previous studies of the Brazilian population. 2,21he association between skin color and health care is complex and appears to refl ect difference in access to health care between the racial strata, 2 motivation in oral health care activities, 22 and the stigma and material privation associated with racial condition. 8come was associated with the prevalence of untreated dental caries, both at the level of the children examined (household income) and at the level of the municipality in which they live (median income).8][9]22 Previous studies of the Brazilian population report low income to be associated with higher levels of intake of sugary foods, 15 less access to health care services 22 and poorer standard of oral hygiene. 5These factors mediate the effect of material privation on the increased risk of dental caries.
The multilevel model meant that the addition of fl uoride to the public water supply could be recognized as a contextual protective factor against the individual risk of dental caries.The negative association between these variables remained signifi cant even after adjusting for individual-level covariates.The benefi ts of fl uoridating the water supply have been well documented in population based studies in Brazil. 17,20A systematic review 25 of the literature concluded that this measure is the most effective and socially acceptable resource in preventing dental caries.
In addition to being a low cost strategy, it may contribute to diminishing health care inequalities between the socioeconomic strata. 6However, the preventative action of fl uoridated tap water has not benefi tted all of the population equally.There is a socioeconomic gradient to the implementation of this measure in different states and regions of Brazil, which contributes to increasing the social bias of this disease. 6,20Therefore, the fi ndings of this study reinforce the need to promote the expansion of fl uoridation to towns that have not yet adopted it.
This study documents current levels of epidemiological indicators of dental caries in deciduous teeth of Brazilian children, as well as the persistent inequalities in suffering from dental caries at regional and socioeconomic levels.These records can be seen as a strong point of the study.It is important to bear in mind that the database used was created through an extensive research effort sponsored by the Brazilian Ministry of Health.It was the first nationwide epidemiological survey of oral health, carrying out home visits in order to examine the children and interview their parents and guardians.There was no chronological difference between the contextual variables and the outcome used in this study, as the information from the most recent general census was collected in the same period in which the data on dental caries were gathered.Moreover, the analysis took into account the complex structure of the cluster sample and the sample weights, issues whose importance has been recognized in epidemiological analysis. 16 is important to highlight that, as they were based on the "d" component (decayed deciduous teeth) of the dmft index, estimates of the prevalence of untreated dental caries do not refl ect the immediate need for dental treatment, in function of the intrinsic characteristics of the instrument.Deciduous incisors with dental caries at age fi ve, for example, are ready to exfoliate and, although they are affected by dental caries, they do not need to be treated.This observation is recognized as a limitation of the study.
This study reinforces the premise that contextual level inequalities continue to infl uence the prevalence of dental caries in pre-school aged children in Brazil.This affi rmation is important in guiding public policies aimed at those geographical areas which are most materially deprived.Reducing inequalities in oral health should be a terget in planning preventative programs which aim to promote both health care and social justice.
The categories used by the Fundação Instituto Brasileiro de Geografi a e Estatística (IBGE, Brazilian Institute of Geography and Statistics) were used to classify skin color; information on household income was collected a Ministério da Saúde.Coordenação de Saúde Bucal da Secretaria de Assistência à Saúde.Projeto SBBrasil 2010 -Pesquisa Nacional de Saúde Bucal.[cited 2013 Sep 04].Available from: http://dab.saude.gov.br/cnsb/sbbrasil/index.html by considering the sum of the income (in reais, R$) of all members of the household.

Figure .
Figure.dmft index and 95% confidence intervals in fi ve-year-olds by macro-region.SBBrasil, 2010.North The dmft index was 2.41 (95%CI 2.19;2.63)for the country as a whole.The "d" component of the dmft corresponded to 84.3% of the total value for the index; this values indicates that, for each 100 teeth with signs of current or previous experience of the disease, around 85 have not been restored.The prevalence of dental caries was 53.1% (95%CI 50.1%;56.1%),in other words, more than half of the children in this age group suffered from the disease.The prevalence of untreated dental caries was 48.2% (95%CI 45.2%;51.2%),suggesting that almost half of Brazilian children in this age group have this condition.

Table 1 .
dmft index, prevalence of dental caries (one or more teeth affected) and the prevalence of untreated dental caries (evaluated using the "d" component of the dmft index) in fi ve-year-old children, by state.SBBrasil, 2010.

Table 2 .
Multilevel analysis of individual and contextual factors associated with the prevalence of dental caries and the prevalence of untreated dental caries (component "d" of the dmft index) in fi ve-year-old children.SBBrasil, 2010.
a Prevalence weighted by sample weight.b Prevalence rates and the 95% confi dence level, unadjusted for other factors, estimates using multilevel analysis Poisson regression.

Table 3 .
Multilevel analysis of individual and contextual factors associated with the prevalence of untreated dental caries in fi ve-year-old children, adjusted model.SBBrasil, 2010.Model 1 does not contain adjustment factors (empty model).Model 2 shows the fi t of the outcome for individual-level variables.Model 3 shows the adjustment for individual-level and contextual variables.b Prevalence ratios and 95% confi dence intervals adjusted for the other factors, estimates using Poisson multilevel regression analysis. a