Oral health profile among community-dwelling elderly and its association with self-rated oral health

REV BRAS EPIDEMIOL 2018; 21(SUPPL 2): E180012.SUPL.2 ABSTRACT: Introduction: The use of dental prosthesis and the tooth loss in elderly people are associated with significant impact on the overall health and quality of life. Continuous assessment of oral health profile in this population is important for planning the actions and policies of the area. Objectives: The aims of this study were to assess the prevalence of tooth loss and use of dental prosthesis among the elderly people in different periods, to evaluate the association between functional dentition (20 teeth or more) and socioeconomic factors, and to evaluate the impact of tooth loss and use of dental prosthesis on self-rated oral health. Methods: The sample consisted of people aged 60 years and older who participated in the Health, Well-Being, and Aging Study (SABE). Data from the years 2000, 2006, and 2010 were used to assess the prevalence of tooth loss and use of dental prosthesis. Analysis of the factors associated with the functional dentition and self-rated oral health was based on the data collected in 2010. Comparison of oral health profile over the 3 years was done through descriptive analysis and comparison of confidence intervals. Multiple logistic regression models were used to assess the factors associated with functional dentition and self-rated oral health. Results: The prevalence of tooth loss and use of dental prosthesis remained constant over the three periods analyzed. Functional dentition was significantly associated with education, sex, and race/gender. Individuals in need of dental prosthesis and with periodontal pocket were more likely to report poor oral health. Conclusion: There was no reduction in the prevalence of tooth loss and in the use of dental prosthesis over 10 years. Functional dentition is associated with socioeconomic inequalities. Self-rated oral health is associated with the need of dental prosthesis.


INTRODUCTION
Prevention and control of diseases and disorders related to oral health are important for maintaining the overall health and quality of life.Use of dentures and tooth loss among the elderly have been linked to overall health 1 and quality of life 1,2 .Sequelae related to oral diseases are among the 50 most common observed in the study on Global Burden of Disease, and edentulism ranks 36th, reaching about 2% of the world population 3 .This condition was ranked 79th among the causes of Disability-Adjusted Life Years (DALY) 4 , despite a reduction seen between 1990 and 2010.Among all oral conditions, this was the leading cause of DALY among people aged 60 years or more 5 .
reduction in the percentage of missing teeth [8][9][10][11] .The high percentage in the use of dental prostheses by the elderly reflects the magnitude of tooth loss and edentulism in later ages 8 .However, despite the high dental impairment, studies show that self-perception of oral health by most individuals is good in this age group, so the role of use and need of prostheses must be elucidated [12][13][14] .
Continuous assessment of the oral health profile of the elderly is necessary for planning appropriate intervention and for assessing the ultimate effect of oral health policies in the population 15 .However, oral health surveys in Brazil have brought information only about people aged 65-74 years, thus leaving a gap when it comes to oral health profile among the population below 65 years and above 74 years of age.
This study was conducted with the following objectives: to assess the prevalence of tooth loss and use of dental prosthesis in a representative sample of noninstitutionalized elderly from São Paulo, Brazil, at different times in 5 year intervals; to evaluate the socioeconomic and demographic characteristics associated with functional dentition; and to evaluate the impact of the use of dental prosthesis and tooth loss on self-rated oral health.

METHODS
We conducted a cross-sectional study based on the data from the Health, Well-Being, and Aging Study (Saúde, Bem-Estar e Envelhecimento -SABE).SABE was initiated in Brazil in 2000 and included a representative sample of people aged 60 years and older living in the urban area of São Paulo.In 2006 and 2010, the second and third waves of the study were conducted, and new cohorts with people aged 60 to 64 were years initiated to maintain the representativeness of this age group in the sample.Participants answered a structured questionnaire and were clinically assessed by dentists trained in a standard methodology 16 .Detailed information of the study is available in other publications 17,18 .

USE OF PROSTHESIS AND TOOTH LOSS
Analysis of oral health status was based on the data collected in 2000, 2006, and 2010.In 2000, only self-report on the number of missing teeth and the use of dentures were evaluated.Self-reported use of dentures was evaluated by the question "Do you use bridge, dentures, or false teeth?" with two possible answers (yes and no).Self-report of tooth loss was assessed by the question "Do you have missing teeth?" with four options: 0-15 (no/a few [up to 4] yes/many [more than 4 and less than half]), and 16 or more (yes, most of them [half or more]).
Self-reported measures in 2000 were compared with tooth loss and the use of prosthesis clinically evaluated in 2006 and 2010.The use of dentures was evaluated for the maxilla and mandible according to the following categories: no, fixed/removable prosthesis, dentures, fixed/removable prosthesis/dentures), which were recategorized simply as yes and no.Number of missing teeth was grouped using the same categories as in the self-report.In 2006 and 2010, the number of missing teeth (edentulous, 1 to 10, 11 to 20, 21 or more) and the time elapsed since the last dental visit were compared.

OUTCOMES
Evaluation of the factors associated with functional dentition (presence of 21 or more teeth) 10 and self-rated oral health was carried out with the data from 2010.
Self-rated oral health was evaluated by the question "How do you rate your oral health?" with five response options (very good, good, regular, poor and very poor), recategorized in good (very good/good) and poor (regular/poor/very poor.) 14

INDEPENDENT VARIABLES
Besides the use of prostheses, the following variables were used to investigate factors associated with self-assessment of oral health in 2010: clinical measures of oral health (number of teeth [0-20 teeth, 21 teeth or more] 10 , need for dental prostheses [yes, no], periodontal pocket ≥ 4 mm [yes, no], periodontal attachment loss ≥ 4 mm [yes, no], need for dental treatment [yes, no], presence of caries (crown or root) [yes, no]); use of dental services (time since last dental appointment [≤ 2 years (< 1 year/1 to 2 years), and 3 years or more (3 years or more/never]); reason for the last dental appointment (urgency, treatment, maintenance); general health (number of self-reported diseases [0-1 disease, 2+ diseases] (including: high blood pressure, diabetes, chronic lung disease, heart disease, osteoarticular disease and stroke); depression 19,20

STATISTICAL ANALYSIS
Oral health status, with regard to tooth loss and use of prostheses in the years 2000, 2006, and 2010, was evaluated by descriptive analysis using frequency measurements and by the comparison of 95% confidence intervals between periods.
The evaluation of the factors associated with self-reported oral health and tooth loss in 2010 was made by multiple logistic regression models.Variables were included in the model in the following order: sociodemographic variables and oral health measures.The analysis was performed in Stata 13.0, using the "survey" command, which allows considering the complex structure of the sample such as the assignment of sample weights.New sample weights were calculated for each wave, thus allowing the maintenance of samples' representativeness.

ETHICAL CONSIDERATIONS
The project SABE was approved by the Ethics Committee of the School of Public Health of Universidade de São Paulo (USP).All selected participants received detailed information about the research objectives, the anonymity guarantee, the forms of data disclosure and the benefits of this study.At the time of interview, researchers and participants signed an informed consent form.

RESULTS
In 2000, 2006, and 2010, the data from 2,143, 1,394, and 1,242 elderly were analyzed, respectively.Because of the incomplete data found for some variables, evaluation of factors associated with tooth loss was conducted in 1,078 cases and analysis related to self-assessment of oral health was made in 1,082 cases.

TOOTH LOSS AND USE OF PROSTHESIS (COMPARISON BETWEEN 2000, 2006, AND 2010)
Most of the elderly had 16 or more missing teeth and used dental prostheses.The comparison between years showed a reduction in the percentage of missing teeth and use of prosthesis, but these were not significant from year to year (Table 1).
Tooth loss and use of prosthesis according to different age groups are shown in Table 2. Comparison between age groups in all the three years demonstrated that despite a reduction in the percentage of missing teeth and the use of dentures in each group, there was an overlap of confidence intervals over years, leading to a nonsignificant reduction.However, analyses of each year showed that the tooth loss significantly increased with age.Regarding

FUNCTIONAL DENTITION IN 2010
Table 3 shows the characteristics of the population in 2010 and the multiple logistic regression models for factors associated with functional dentition.It was found that 18.6% (95%CI 15.3 -22.5) of the elderly had functional dentition.Most seniors were female, white, and had reported sufficient income for basic expenses.
The chance of having functional dentition was higher among the male and among the well-educated.People aged 75 years or older had less chances of presenting functional dentition compared with those aged 60 to 64 years.People who declared themselves as "black" and "mulatto/brown" were 65 and 45% less likely to have functional dentition.

SELF-RATED ORAL HEALTH IN 2010
Regarding self-rated oral oral health, in the bivariate analysis, it was significantly associated with one sociodemographic variable (sufficient income), three general health variables (number of diseases, depression, and self-rated general health), and all oral health variables, with the exception of functional dentition.
The final logistic regression model showed that people aged 80 years or more and with marital relationship had significantly less chance of self-rated poor oral health.The chance of poor self-rated was higher among seniors with two or more diseases, in need of prostheses, presenting periodontal pocket, and who had visited their dentists 3 years ago or more (Table 4).

DISCUSSION
The main result of this study is the maintenance of the prevalence of tooth loss and use of dental prosthesis over 10 years among the noninstitutionalized elderly of São Paulo.In addition, the prevalence of seniors with functional dentition is low, but higher than the rate observed for the whole Brazilian population 9 and lower than the rates found in developed countries 21,22 .Need for dental prosthesis and the presence of periodontal pockets are significantly associated with poor self-rated oral health.
Different categorizations for tooth loss in different studies make it difficult to compare them.However, unlike in some developed countries [23][24][25] , the stability of prevalence rates among the elderly in Brazil's largest city reproduces the findings of the population in the first decade of the 21st century 26 .
This rate stability could be attributed to the short time of observation, since a large proportion of the elderly participants belong to generations who were not exposed, when young, to the socioeconomic changes and to the prevention and treatment policies that have occurred throughout the country in the past decades.These events have produced significant effect among Brazilian adults 27 .The relevance of the cohort effect in improving oral health status has been highlighted in the literature [26][27][28] .According to Slade et al. 28 , the major determinant of the drop in  the prevalence of tooth loss in the United States was the passage of generations born before the 1940s, whose incidence (5-6% per decade of age) exceeded the rates of later cohorts (1-3% per decade of age).In Brazil, the reduction of edentulism as an effect of the introduction of new cohorts is planned for the decade 2050 26 .
The cohort effect cannot be confirmed by means of the analysis performed in this study; however, this effect may be suggested from the comparison of tooth loss among cohorts of 60 to 64 years (in relation to the elderly of the same age in different years of study), which shows an approximate absolute reduction of seven percentage points in the prevalence of tooth loss in the youngest generation (evaluated in 2010) as compared with the other cohorts.The comparison of cohort of 60 to 64 years in 2006 and 2010 shows that edentulism had an absolute reduction of 6.6 points and an increase of 4.0 points in the prevalence of functional dentition.Regarding dentures, as noted by other authors 23 , the effects of tooth loss reduction are reflected as a decline in the use of dentures and increase in the use of fixed or removable dentures.
With regard to factors associated with functional dentition, after adjusting for overall health variables and use of dental services, the presence of functional dentition was still significantly associated with socioeconomic and demographic variables, which confirms the inequalities observed in other studies 21,24,25,29 .Regardless of the variables used, number of missing teeth, edentulism or functional dentition, the chance of this outcome has been significantly higher among less educated individuals 21,26,29 , where one can say that there is a social gradient in tooth loss 24 .
As for the association with the skin color and gender, results are conflicting.In the Brazilian population aged 65-74 years, skin color and gender have lost significance in association with tooth loss when adjusted for socioeconomic variables 26 .Kida et al. 30 found that Tanzanian women were significantly more likely to present tooth loss than men.Wu et al. 25 found that black Americans had less chances of losing teeth, but a larger number of missing teeth.In the same study, no association between gender and edentulism could be established, but women had less tooth lost over time.
Tooth loss reflects not only the history of oral diseases of individuals, but also the behavior and attitude of the patients and dentists, supply of and access to dental care, and the philosophy of dental care 24 .Some researchers have considered edentulism as a social phenomenon.Among the aspects highlighted, the way each social formation accepts and encourages tooth extraction as an inevitable event, as well as the guidelines for dental services organization are important determinants 31,32 .Thus, edentulism acquired at early ages is related to cultural issues and to limited access to care, equally affecting population groups when it comes to skin color and gender.Functional dentition would be more sensible to socioeconomic differences between population groups which end up conditioning the use of health services.Regarding gender, most tooth loss cases among women may be attributed to the hypothesis that they use more dental services and are therefore more prone to overtreatment 26,30 .
In addition to sociodemographic conditions, functional dentition was associated with smoking and the use of dental services.Not smoking and having made recent use of dental services may reflect greater health care, which would explain the best result of oral health among seniors who were nonsmokers and those whose last dental appointment had been 2 years or earlier.Other studies 21,[33][34][35] have reported association between smoking and tooth loss, being the destruction of dental-supporting tissues caused by tobacco consumption pointed out as one of the ways by which smoking would be associated with tooth loss 34 .
Regarding self-reported oral health, after adjusting for sociodemographic, general health and oral health variables, older people in need of prosthesis and presenting periodontal pockets were more likely to self-report poor oral health, which shows the importance of public dental services ensuring dental care and provision of dental prosthesis for this population.As noted by other authors, no association was found between this outcome and the functional dentition 12,13,36 and the use of prostheses 12,13 .These findings indicate that good oral health evaluation may be achieved in the presence of functional dentition or use of dental prostheses, as the need for prosthesis related to poor self-assessment represents the absence of such conditions.
Among the strengths of this study, it is worth highlighting the use of data from a representative sample of noninstitutionalized elderly residents in the largest city in South America, where the proportion of seniors, altogether, is quite significant.Among the limitations, one can mention the use of self-reported dental conditions in the first year of the study, which may have generated some bias in estimates.However, several studies have demonstrated the validity of self-reported number of teeth [37][38][39] and use of prostheses 37,40 in epidemiological studies.Another limitation was the impossibility of establishing causal relationships between outcomes and covariates, given the cross-sectional design of the study.

CONCLUSION
Based on the foregoing, the three surveys showed that the prevalence of both tooth loss and use of prosthesis was high and remained constant.As per the study held in 2010, we can confirm the existence of social inequalities with respect to functional dentition.Moreover, the need for prosthesis and the presence of periodontal pockets were associated with self-rated oral health.These results reinforce the importance of oral health services planning throughout life, being continuously evaluated, with the perspective of changing, for next generations, this unfavorable profile of oral health found in the elderly population of this study.

Table 1 .
Distribution of oral health status of the elderly in 2000, 2006, and 2010.

Table 2 .
Distribution of oral health conditions of the elderly according to age groups(2000, 2006, and 2010).

Table 3 .
Distribution of study variables and multiple logistic regression model for the presence of functional dentition.

Table 4 .
Final logistic regression model for poor self-rated oral health among elderly in São Paulo, 2010.