Self-perception of oral health and associated factors in quilombola older people: a population-based study

Objective: to investigate self-perception of oral conditions and associated factors in rural quilombola older people in northern Minas Gerais, Brazil. Method : This is an analytical and cross-sectional population-based study, in which cluster sampling with probability proportional to size (n=406) was used. Data collection involved conducting interviews and clinical dental examinations. Self-perception of oral conditions was assessed using the GOHAI (Geriatric Oral Health Determination Index). Results: Most older people self-perceived oral health as excellent (46.3%) or regular (30.2%). Those surveyed also revealed precarious oral health and restricted access to dental services. In the multiple analysis, significant associations ( p ˂ 0.05) were found between regular GOHAI and variables related to the location of the last consultation and use of prosthesis, as well as between poor GOHAI and variables related to marital status, religion, reason for last consultation, CPO-D index and use of prosthesis. Conclusion: A significant portion of the quilombola older people showed a more positive self-assessment of oral health, which differs from the professionally verified dental condition. It was also found that the report of poor perception of oral conditions was strongly associated with poorer oral health among those investigated.


INTRODUCTION
In the field of public health, there is a clear situation of exclusion and marginality experienced by black and brown people [1][2][3][4][5] .In this context, the remnants of quilombola communities 1,3,4 are included, defined as ethnic-racial groups, holders of black ancestry associated with resistance resulting from historical oppression6.These peoples generally manifest a rural geographic distribution 3,4,7 , totaling 3212 quilombola communities certified by the Palmares Foundation throughout the Brazilian territory 8 .
It is useful to emphasize that oral health directly impacts people's quality of life 9 .Specifically in older people, certain oral alterations are prevalent and become objectively relevant, such as edentulism, dental caries, periodontal disease, need for prostheses and lesions of the oral mucosa 3,10,11 .In the case of older people residing in rural environments, such as the Quilombolas, the epidemiological scenario is accentuated due to limited access to dental services, thus showing people with a high prevalence of edentulism and oral pathologies 3,5 .
In view of the above and considering the implications of the subjective aspect on people's health, it is emphasized that self-perception measures, in a relevant way, the individual's state of health, since it incorporates cognitive and emotional aspects, as well as physical 12 .Complementarily, it is highlighted that it is essential to understand how people perceive their oral condition, since their behavior is modulated according to the relevance attributed to their own oral health 13 .
In older people, the perception of the oral condition can be influenced by personal values, such as the conviction that painful and disabling events are inevitable at this age, thus leading to an overestimation of their condition 11,14,15 .In addition, it is observed, among these individuals, that the lack of perception of their real dental situation results in non-demand for professional assistance 14,16 .
Studies indicate that self-perception of oral health is associated with clinical dental factors 9,11,17 , subjective factors 11 , in addition to being influenced by socioeconomic factors such as age and income 18 .
However, it appears that there is an intense shortage of studies in the literature that discuss the self-perception of oral conditions in quilombolas, primarily covering older people.The sparse data available reveal individuals with poor oral health and negative self-perception of oral conditions 4,19 .Thus, it is inferred that the epidemiological data expressed by this investigation can encourage the proficient elaboration of public dental policies aimed at the local public.
Thus, the objective of this research was to investigate the self-perception of oral conditions and associated factors in rural quilombola older people in the north of the state of Minas Gerais, Brazil.

METHODOLOGY
The detailed methodological description of this study regarding its design, setting 20 , population universe 8,21,22 , sampling plan, inclusion 8 and exclusion 23 criteria, data collection and ethical aspects, is fully included in a previous related publication 3 .
It should be noted that this is a cross-sectional population-based study, which was carried out in the extension of the northern health macro-region, located in northern Minas Gerais, Brazil.The region had 79 quilombola communities, encompassing about 19,000 inhabitants and an older population estimated at 2,660 people (N) 8,21,22 , which totaled a minimum sample calculated for the study of 406 older people (n).For sample calculation purposes, a prevalence of 50% for oral diseases was estimated in a finite population, a confidence level of 90%, margin of error of 5%, design effect (deff ) equal to 1.5 and estimated 10% losses.
Regarding sample selection, cluster sampling with probability proportional to size (PPS) was adopted, thus selecting 30 local communities.The selection process of residences in each community started from a previous definition of the central community region, with subsequent displacement of the researchers in loco in a spiral direction (considering the prevalent geographical distribution spaced between residences in these rural communities), going through households and carrying out examinations and interviews, until reaching the sample defined for each community.All older people (≥ 60 years old) in the households were invited to participate in the survey.
Inclusion criteria were established to be at least 60 years old, self-declare as a remnant of quilombo and reside in a quilombola community certified by the Palmares Cultural Foundation 8 .People who showed cognitive deficits were excluded from the research, a condition that could make it difficult or impossible to transmit information regarding the researched variables.Cognitive deficit screening was performed using the Portuguese version of the Mini-Mental State Examination (MMSE), translated and modified 23 .
The dependent variable in this study was selfperception of oral health and the independent variables related to sociodemographic characteristics, access to dental services and oral condition.The dependent variable was evaluated through an interview, using the Geriatric Oral Health Determination Index -GOHAI, composed of 12 questions.Each question had three possible answers: "always", "sometimes" and "never".They received scores 1, 2 and 3, respectively, according to each answer, and to determine the global index, the score of each question was added.Each individual's score could range from 12 to 36 13 .Thus, the individual's selfperception was classified as "excellent" (34 to 36 points), "regular" (30 to 33 points) and "poor" (< 30 points) 24,25 .The prevalence of dental caries was investigated using the DMFT Index, considering the number of decayed (D), missing (M) and filled (F) teeth, and the periodontal condition was assessed using the Community Periodontal Index (CPI) and Clinical Attachment Loss (CAL) 26 .
Data tabulation and analysis were performed using statistical software.Initially, a descriptive analysis of the data was performed.Subsequently, a bivariate analysis was conducted to verify the association between self-perception and variables concerning sociodemographic characteristics, access to dental services and oral conditions, using Pearson's chi-square test -X².Finally, a multiple analysis was performed, adopting the Multinomial Logistic Regression Model and using the variables that presented a value of p≤0.25 in the bivariate analysis.The category adopted as a reference for the dependent variable was GOHAI excellent.In the multiple analysis, the hierarchical method 1,27 was adopted, as shown in Figure 1, for entering the variables in the model, using the following sequence: distal level (sociodemographic variables), intermediate level (access to oral health services) and proximal level (oral health conditions).Within each hierarchical level, for selection of variables, the backward selection method based on the Wald test was adopted, thus all independent variables were added to the model and the removal of these variables was made based on the significance of the Wald test.The adequacy of the model adopted in the multiple analysis was verified using the Likelihood Ratio test and the quality of the model's fit was tested using the Deviance test.Furthermore, the Pseudo-R 2 Nagelkerke was used in order to obtain the explanatory power of the final model.After regression, in the final adjusted model, the magnitude of association between variables was estimated using the odds ratio (OR) and the α significance level considered was 5%.
This study followed the precepts enshrined in Resolution No. 466/2012 of the National Health Council of the Ministry of Health and those dictated by Resolution CFO 179/91 of the Code of Professional Dental Ethics.Approval of this research was carried out by the research ethics committee of the State University of Montes Claros (COEP-Unimontes), through consolidated opinion nº 2,821,454.Participants were informed about the research, as well as instructed to sign the free and informed consent form for participatory permission and data analysis.

RESULTS
Figure 2 shows the relative frequency of categories related to self-perception of oral health in the investigated quilombola older people (n=406).There was a predominance of self-perception classified as excellent (n=189; 46.3%; 95% CI 40.5 -52.2), followed by regular self-assessment (n=123; 30.2%; 95% CI 25.3 -35.6) and poor (n=94; 23.5%; 95% CI 18.2 -29.8).It is pointed out that the expressed confidence interval (CI) was corrected by the effect of the study design and that the descriptive data of the sample are largely based on a previous related publication 3 .Table 1 expresses the bivariate analysis of the data, revealing the statistical association between the dependent variable (GOHAI) and the independent variables.In this phase, the variables marital status, skin color, reason for last consultation, evaluation of care, use of prosthesis, need for prosthesis, DMFT, CPI, CAL and soft tissue alteration were associated with GOHAI, at a significance level of 25%.It should be noted that the variable "Where was the last consultation" originally manifested the category "never been to the dentist" containing only nine respondents, thus making the multiple analysis unfeasible and requiring, therefore, an adequacy of the referred variable with the removal from the aforementioned category.Thus, after adjustment, the variable "Where was the last consultation" presented, in the bivariate analysis, a value of p=0.065, revealing an association with the dependent variable and, therefore, was included in the multiple analysis.
Table 2 demonstrates the Mutinomial Multiple Logistic Regression Model, showing the existing statistical associations between the dependent variable (GOHAI) and independent variables that were included in the final model.At this stage, there was a statistically significant association ( p˂0.05) between regular GOHAI and the variables related to the place of the last appointment and use of prosthesis, as well as between poor GOHAI and variables concerning marital status, religion, reason for last appointment, DMFT and use of prosthesis.
Regarding the adequacy of the model used in the multiple analysis, a significance test of the adjusted model was carried out, using the Likelihood Ratio test (χ 2 (16) = 60.97;p<0.0001), indicating that the general model had at least one independent variable associated with the outcome.The model's goodness of fit test was performed using the Deviance test (χ 2 (148) =150.25;p=0.433), indicating that the model fitted the data observed in the sample.Finally, the explanatory power of the final model was obtained through Pseudo-R 2 Nagelkerke = 0.166 (16.6%), representing the proportion of information gain estimated by the full model compared to the null model (the closer to 1, the better).It is added that the partial explanatory power of the model with the variables of the distal level was Pseudo-R 2 Nagelkerke = 0.04 (4%); with the variables of the distal and intermediate levels, the partial explanatory power was Pseudo-R 2 Nagelkerke = 0.104 (10.4%).

DISCUSSION
The data from this study revealed that a significant portion of the surveyed quilombola older people (46.3%) self-perceived their oral health as excellent and that only 23.5% of those investigated reported a poor self-perception.Thus, there was an evident divergence between the clinical dental condition objectively measured and the self-reported oral health by the respondents.Similar results were found in older quilombola people in southern Brazil and Minas Gerais 4,7,19 .It is inferred, therefore, that diverse cultural aspects, combined with the social inequalities experienced by quilombola older people, including the severe deficit in access to public oral health services, can lead them to a state of longitudinal resignation, culminating in acceptance of their adverse dental condition and consequent manifestation of self-assessment correlated with a more positive trend.
An indictable dissonance was also found among Brazilian older people, through national surveys 11,28 , and would be linked, according to the authors, to personal and cultural issues 11,14,15 .It is clear, therefore, that there is a tendency for people in this age group to overestimate their oral condition, even in the face of a compromised dental situation 11,14,15 .Complementarily, it is emphasized that the professional evaluates the dental clinical condition according to the presence or absence of diseases, while for the patient, the symptoms and functional and social consequences resulting from oral diseases become relevant 13 .
Following the opposite trend, a study conducted with urban and rural northeastern older people indicated that residents in the rural area had worse oral conditions and a higher prevalence of negative self-perception of oral health, when compared to those who lived in the urban area 29 .Research involving rural and urban older people, from a significant Chinese province, showed that 35.1% of the participants self-perceived oral health as poor, coinciding with the precarious oral condition clinically observed by the researchers 30 .Punctuated convergence between the dental clinical picture and self-perception of oral health can signal and ratify the relevance and impact that socioeconomic, personal and cultural issues can assume in diverse rural populations, greatly interfering in the self-report of oral conditions.
With regard to the multiple analysis carried out in this study, it was found that poor self-perception of oral conditions was associated with having a partner, being evangelical, having a high DMFT index, not using a prosthesis and reporting dental pain as reason to seek professional assistance.Consequently, individuals who had a partner were more likely to have a poor self-perception of oral health.Corroborating this finding, Miranda et al. 17 demonstrated, in a multiple analysis, that older people who did not have a partner were more likely to have a more positive self-assessment of oral health.In both studies, the authors did not show a plausible explanation for the aforementioned association, and the same thing happened with the religious issue observed in this research.
Dental pain as a motivating factor for seeking care proved to be an important variable related to poor self-perception of oral health, with a significant odds ratio.Martins et al. 16 found a lower prevalence of positive self-perception of oral health among older people who sought dental services when faced with the perception of an oral problem.Still in this scope, dental pain was also associated with negative self-assessment of oral health among rural adults in the Brazilian Northeast 31 , corroborating the strict correlation between these variables.Similarly, a recent longitudinal study demonstrated that the worsening of self-rated health in older people is strongly associated with the presence of chronic diseases and the manifestation of frailty, confirming the impact of illness on the decline in self-rated health in individuals in this age group 32 .
In the present study, older quilombola people who had DMFT between 21 and 32 showed a greater probability of self-perceiving their oral health as poor, which is therefore consistent with the professional assessment, emphasizing that most of those surveyed were edentulous.Similarly, a Chinese study, including rural and urban older people, demonstrated that individuals with DMFT ≥ 20 had worse selfperception 30 .Discordant results were found in a national study involving individuals in the express age range, demonstrating that the smaller the number of teeth present in the investigated oral cavity, the greater their positive perception of oral health 11 .Other studies showed that edentulous older people were more likely to self-report their oral health as positive 16,18 .As argued by the researchers, this trend could be related to possible negative experiences experienced by people with natural teeth throughout their lives, such as pain or discomfort, thus boosting a better self-perception of oral health in the face of missing teeth 16 .
Quilombola older people who did not use dental prostheses were more likely to have regular and poor self-assessment of oral health, thus showing a strong association between the variables and confirming the impact of missing prostheses on oral selfperception, leading to a more negative self-report 33 .Such data reiterate the need to expand access to oral rehabilitation aimed at quilombola older people, who ordinarily manifest a high prevalence of caries and edentulism 34 .
Focusing on the hierarchical model adopted in this investigation, it was found that the explanatory power of the final model was 16.6% in relation to the variability of the dependent variable (selfperception of oral health), and the independent variables that most contributed to the outcome were those concerning access to dental services (6.4%).Thus, it became evident, among the quilombola individuals surveyed, the primary influence that the restricted access to dental services had on the relative variation in the self-assessment of oral conditions, thus confirming the pressing local need to expand public services of dental nature.
Silva and Castellanos-Fernandes 13 , when researching older people from São Paulo, showed that the independent variables used explained 30% of the variation related to self-assessment in oral health, with a greater contribution from clinical dental variables related to DMFT (recommended extraction) and CPI (community periodontal index), which accounted for approximately 1/3 of the variability.Among Brazilian older people, the model used in the study was able to explain 50% of the variability in self-perception of oral health among dentate individuals and 43% among edentulous individuals, with a greater contribution from subjective conditions related to oral health, such as the appearance of teeth, gums and mastication 11 .It appears, therefore, that there is a relative variation in the contribution exerted by different independent variables on the variability of self-perception in oral health, a volubility that may be associated with specific social and cultural aspects, therefore requiring further investigations aimed at a fruitful scientific clarification.
It should be added that the data obtained in this investigation have external validity, being, therefore, fully extendable to the quilombola older people residing in the northern health macro-region of the state of Minas Gerais.It is noteworthy that geographic access to local quilombola communities proved to be a prominent obstacle in the course of this study.It is also reinforced that this research does not allow inferences regarding causality or temporality, since it is a cross-sectional study.Furthermore, it is pointed out that the self-declared information emanating from this study is sensitive to possible memory biases coming from the participants, which can interfere with the accuracy of the data collected.Thus, in view of the patent scarcity of research involving topics concerning the self-perception of oral conditions in older quilombola people, it is suggested that new national investigations be carried out, with a proficient purpose directed to diagnosis and planning in oral health, to scientific parameterization and to the improvement of the quality of life of this population.Finally, the deleterious impact that oral pathologies can have on quilombola older people is highlighted, as observed, which may result in a selfperception of oral conditions with a more negative tendency.Express scientific evidence signals and encourages the effective implementation of public policies at the local level, with a view to strengthening community dental care, thus contributing to the well-being of these people.

CONCLUSION
It was verified that the majority of older quilombola people expressed an excellent or regular self-perception of oral health, different from the objective diagnosis verified professionally.In the final statistical model adopted, the self-assessment of oral health was associated with the independent variables related to the place and reason for the last dental visit, use of dental prosthesis, marital status, religion and DMFT index.Furthermore, the findings of this study demonstrate an evident local deficit of accessibility and completeness of care of public oral health services, which, according to national guidelines, must prioritize the family focus and community orientation, with the aim of guaranteeing equity in health.Additionally, it is emphasized that this study provides scientific subsidies so that local communities can, through social control of the health sector, claim improvements in dental care and a longitudinal planning of public policies that is oriented to the urgent needs identified.

Figure 1 .
Figure 1.Hierarchical conceptual model, structured in variable input levels, used in the multiple analysis (dependent variable: self-perception of oral health).

Figure 2 .
Figure 2. Distribution of self-perception of oral health in rural quilombola older people in northern Minas Gerais (n=406), Brazil, 2019.

Table 1 .
Bivariate analysis of data relating to rural quilombola older people in northern Minas Gerais (n=406), Brazil, 2019.
* The reference category of the dependent variable was GOHAI Excellent.OR: odds ratio; CI: confidence interval;