Risk indicators for tooth loss in Kiriri Adult Indians: a cross-sectional study

Abstract

Background: The aim of this cross-sectional study was to evaluate the risk indicators of tooth loss in adult Kiriri Indians from Brazil. Methods: A representative sample of 225 Indians (≥19 years of age) was assessed. Interviews using a structured written questionnaire were performed to collect data on demographics and socio-economic status, and health-related data. Probing depth, the distance between the cement–enamel junction and the free gingival margin, and decayed, missing or filled teeth were evaluated. Bivariate and logistic models were used to assess associations between tooth loss and age, sex, income, education, diabetic status, smoking habits, dental caries, severe periodontitis, plaque index and previous dental visit. Results: Eighty per cent of subjects had lost one tooth or more, and 20% had lost eight teeth or more. Mean (±standard deviation) tooth loss was 5.09 (±5.83) teeth. After adjustment for covariates, loss of one tooth or more was associated with older age [≥35 years; odds ratio (OR) = 4.06, 95% confidence interval (95% CI): 1.38–11.94, P = 0.01], severe periodontitis (OR = 3.35, 95% CI: 0.99–11.24, P = 0.05), higher dental caries (OR = 3.24, 95% CI: 1.35–7.78, P = 0.01) and previous dental visit (OR = 23.32, 95% CI: 5.75–94.63, P < 0.001). Conclusion: Tooth loss is highly prevalent in Kiriri Indians. Older age, severe periodontitis, higher caries index and previous dental visit were associated with tooth loss. Prevention and treatment programmes, targeting high-risk groups, are required to promote the oral health of the population.

INTRODUCTION

Tooth loss is a measure of a population’s oral health status. Studies of trends of tooth loss in different populations provide important information about the disease burden, risk factors for tooth loss, potential changes in oral health status, possible causes of these changes and treatment needs1., 2., 3., 4..

Tooth loss has decreased significantly in many populations during recent decades, mostly because of water fluoridation, use of fluoride toothpaste and community-based prevention programmes in public dental health services2., 3., 4., 5., 6., 7., as well as increased awareness of the importance of oral health in the population⁸. However, tooth loss is still highly prevalent. Forty-three per cent of American adults, 65 years of age and over, have lost six or more teeth⁹. In Brazil, Canada, Portugal and Turkey, the prevalence of edentulism is over 50%¹⁰. In Brazil, mean loss of 13.5 teeth has been described among younger adults (35–44 years of age)¹¹. Low social class or income, low education and the presence of dental caries and periodontal diseases have been associated with tooth loss12., 13..

Inequalities in unmet dental needs are particularly evident in Indian groups¹⁴ and in other racial and/or ethnic minorities¹⁵. However, knowledge of tooth-loss frequency and its risk factors in Indian populations is scarce because of logistical, financial and political constraints¹⁶. A single study evaluated tooth loss in 12,349 American Indian and Alaska Native adults. Complete tooth loss in Indians ≥35 years of age was 11%; and in Indians ≥65 years of age, it was 42%. The mean number of remaining teeth in Indians ≥35 years of age was 20.7. Only 20% of subjects between 35 and 44 years of age had not lost at least one tooth. Between 1984 and 1991, the rate of people with 20 or more teeth increased. Tooth loss was more severe in subjects with diabetes¹⁷. Studies on other adult Indian populations showed that mean tooth loss was between 0.4 and 13.7 teeth per person, depending on the age group18., 19., 20..

Kiriri Indians are indigenous people of Brazil. At the time of this writing, the Brazilian government recognised 240 Indian tribes, who comprised around 900,000 people (or 0.4% of Brazil’s population). The Kiriri Indians live in north-east Brazil and comprise an isolated population of approximately 2,182¹⁶. Access to the area inhabited by this population is challenging, and external influences are limited, as shown by maintenance of the social, cultural and behavioural backgrounds of the population¹⁶. Health-care traditions using plant-based medicines and prayers, lack of help with transport for patients from their remote geographical location, poor quality of the public health-care service and high costs of private health care inhibit access to dental services¹⁶. Previous studies in this population have shown a prevalence of destructive periodontal disease of 97.8% and a prevalence of mucosal lesions of 22.4%16., 21.. The aim of this study was to evaluate risk indicators of tooth loss in adult Brazilian Native Kiriri Indians.

MATERIALS AND METHODS

The present study is part of a larger project, and some of the results related to periodontal diseases and mucosal lesions have been previously published16., 21.. The study was performed in accordance with the World Medical Association Declaration of Helsinki and was approved by the Brazilian Research Ethics Committee of the Ministry of Health, Brasilia, Brazil, and by Indian authorities. Subjects who agreed to participate signed an informed consent form.

Study design

This study was a cross-sectional survey. The target population was Kiriri Indians, ≥ 19 years of age, who were living in an isolated Indian area in Bahia state in north-east Brazil.

The study sample was drawn from a sample of subjects, ≥19 years of age, among the Kiriri Indian population, from a study of periodontal health¹⁶. A representative sample of adults was calculated for the periodontal health study from a list with the names of all Kiriri Indians provided by the Brazilian National Health Foundation. As previously described, of the 2,182 Kiriri Indians living in the Kiriri Indian region in 2011, 1,025 were adults (≥19 years of age), and the individuals were randomly selected for study. In brief, a sampling error of 5%, a confidence level of 95% and maximum percentage of periodontal disease of 79% were considered. The calculated sample was 205 individuals. Considering a response rate of 90%, 226 individuals were invited to participate¹⁶. The response rate was 99.6%, and 225 individuals between 19 and 77 years of age were evaluated (Table 1).

Table 1.

General characteristics of the sample (Kiriri Indians, Brazil, 2011; n = 225)

Variable	Total sample
Age (years)
19–34	137 (60.9)
≥35	88 (39.1)
Education
≥9 years	156 (30.0)
<9 years	67 (70.0)
Income
≥US$259.00	36 (16.1)
<US$259.00	187 (83.9)
Smoking habits
No	111 (49.3)
Yes	114 (50.7)
Diabetes
No	208 (93.7)
Yes	14 (6.3)
Dental caries
≤3	129 (57.3)
>3	96 (42.7)
Destructive periodontal disease
No	135 (60.0)
Yes	90 (40.0)
Plaque index
<40%	140 (62.2)
≥40%	85 (37.8)
Previous dental visit
Yes	204 (91.0)
No	21 (9.0)
Mean percentage of tooth loss/subject	5.09 ± 5.83
19–34 years of age	2.47 ± 2.69
35–44 years of age	6.44 ± 5.66
≥45 years of age	11.57 ± 7.15

Values are given as n (%).

Operational procedure in the field

First, the research project was presented to the local Indian authorities, who invited the community to participate. Second, the community dentist and two nurses visited the selected subjects to explain the aims of the project and to encourage participation. Before oral examination, two Indian nurses conducted in-person interviews using a structured written questionnaire to collect data on demographics, socio-economic status and health-related issues. After the interview, the same nurses performed a fasting glucose blood test (OneTouch Ultra Mini; Lifescan, Milpitas, CA, USA) of all the selected Indians¹⁶.

Oral health evaluation

Four trained dentists, assisted by four trained undergraduate students from the Dentistry School of the Federal University of Bahia, performed the oral examinations. The oral examinations were conducted using a headlight (Turboled; Nautika, São Paulo, São Paulo, Brazil), with the subjects seated on a standard chair used in schools and health-care facilities¹⁶.

Before the study, the examiners were calibrated for accuracy and repeatability using a group of 10 subjects. The intraclass and interclass correlation coefficient (ICC) values revealed reproducibility according to site level for probing depth (PD), the distance between the cement–enamel junction and the free gingival margin (CEJ-GM), and identification of decayed, missing or filled teeth (DMFT) (intra-examiner ICC was ≥0.81, and interexaminer ICC was ≥0.75).

All permanent teeth, excluding the third molars, were examined. Tooth surfaces were scored as sound, decayed (D), missing (M) or filled (F) teeth according to the World Health Organization (WHO) criteria²². A tooth was diagnosed as sound when there was no evidence of restorative treatment or untreated dental caries. A tooth was considered as decayed if there was ‘cavitation of enamel or dentinal involvement or both being present’ or ‘visible caries that are contiguous with a restoration’, as well as total/partial coronal destruction resulting from disease evolution. Filled was recorded when a tooth contained one or more permanent restorations. Missing teeth included teeth that had been extracted for decay or for other periodontal reasons, congenitally missing teeth and unerupted teeth. In cases where there was uncertainty about the diagnosis, the tooth was classified as sound.

Periodontal evaluation was performed using a manual periodontal probe (PCP-UNC 15; Hu-Friedy, Chicago, IL, USA). PD and the CEJ-GM were measured at six sites per tooth (mesiobuccal, midbuccal, distobuccal, distolingual, midlingual and mesiolingual). Measurements were made in millimeters and rounded up to the next whole millimeter. Clinical attachment level (CAL) was calculated as the sum of the PD and CEJ-GM parameters¹⁶.

Data analysis

The statistical analysis included 225 subjects. Tooth loss was categorised as loss of one tooth or more, or no tooth loss. Age, education, socio-economic status and diabetes were categorised as previously described¹⁶. Age was categorised as 19–34 or ≥35 years of age for analysis of the prevalence and extent of tooth loss. Education level was categorised as ≥9 years of education or <9 years of education. Socio-economic status was categorised as monthly income <US$259.00 or ≥US$259.00. Subjects were classified as current non-smokers or smokers. Subjects were classified as diabetic based on self-reported physician’s diagnosis of diabetes, or a fasting blood glucose level of ≥126 mg/dL, associated with diabetes symptoms (increased thirst, increased urination and unexplained weight loss). If neither of these two conditions was met, subjects were considered non-diabetic. Severe destructive periodontal disease was defined as subjects with two or more proximal sites with CAL ≥6 mm, not on the same tooth, and one or more proximal site(s) with PD ≥5 mm²³

Descriptive statistics of tooth loss were calculated according to age, sex, education, income, smoking status, diabetic status, dental caries, severe periodontitis, plaque index and previous dental visit. The chi-square test was used to compare subjects with and without tooth loss, according to the aforementioned variables.

Backward stepwise logistic regression was used to determine the variables associated with tooth loss. Odds ratios (ORs) and 95% confidence interval (95% CI) were calculated. The chosen level of significance was 5%.

Data analysis was performed using a statistical software program (SPSS version 13.0; SPSS Inc., Chicago, IL, USA).

RESULTS

Sixty-one per cent of subjects were women. Most of the subjects were young adults <35 years of age (61%), with <9 years of education (70%), and an income lower than US$259.00 per month (84%). Only 6.3% of subjects were diabetic. Fifty per cent were smokers.

The median number of caries lesions was three, but 43% of subjects had more than three lesions. Forty per cent of subjects had destructive periodontal disease, and 38% had plaque index ≥40%. Only 9% had not visited a dentist before. Eighty per cent of subjects had lost one or more teeth. Mean percentage tooth loss was 5.09 (±5.83) (Table 1).

Bivariate analyses showed that the loss of one tooth or more was associated with older age, lower education, smoking habits, diagnosis of severe periodontitis, higher dental caries rate and a previous dental visit (Table 2). The loss of one tooth or more was not associated with sex, income, diabetes or plaque index. Regression analysis showed that the loss of one tooth or more was associated with older age (≥35 years), higher dental caries experience, severe periodontitis and previous dental visit (Table 3). Sex, income, education level, diabetes, smoking habit and plaque index were not associated with the loss of one tooth or more by regression analysis.

Table 2.

Bivariate analysis of the association of age, sex, income, education, smoking habits, diabetes, dental caries, dental plaque index, destructive periodontal disease and previous dental visit with the loss of one tooth or more (n = 225)

Variable	1 Tooth loss (%)		OR (95% CI)	P value
	No	Yes
Age (years)
19–34	27.0	73.0	4.28 (1.81–10.11)	<0.001
≥35	8.0	92.0
Sex
Female	22.6	77.4	1.55 (0.79–3.06)	0.21
Male	15.8	84.2
Education
≥9 years	28.4	71.6	2.18 (1.09–4.33)	0.02
<9 years	15.4	84.6
Income
≥US$259.00	19.4	80.6	1.01 (0.41–2.49)	0.98
<US$259.00	19.3	80.7
Smoking habits
No	25.2	74.8	2.07 (1.05–4.08)	0.03
Yes	14.0	86.0
Diabetes
No	19.7	80.3	0.90 (0.24–3.38)	0.88
Yes	21.4	78.6
Dental caries
≤3	24.8	75.2	2.31 (1.12–4.77)	0.02
>3	12.5	87.5
Severe periodontitis
No	24.4	75.6	3.87 (1.45–10.32)	0.01
Yes	7.7	92.3
Plaque index
<40%	20.7	79.3	1.22 (0.61–2.43)	0.57
≥40%	17.6	82.4
Previous dental visit
Yes	15.9	84.1	6.46 (2.48–16.83)	<0.001
No	55.0	45.0

95% CI, 95% confidence interval; OR, odds ratio.

Table 3.

Logistic regression for age, dental caries, destructive periodontitis and previous dental visit with the loss of one tooth or more (n = 225)

Variables	OR (95% CI)	P value
Age (years)
19–34	1	0.01
≥35	4.06 (1.38–11.94)
Dental caries
≤3	1	0.01
>3	3.24 (1.35–7.78)
Severe periodontitis
No	1	0.05
Yes	3.35 (0.99–11.24)
Dental visit
Yes	23.32 (5.75–94.63)	<0.001
No	1

95% CI, 95% confidence interval; OR, odds ratio.

DISCUSSION

In the present study, the prevalence of tooth loss and its risk indicators were evaluated in adult Kiriri Indians. The results showed high prevalences and extent of tooth loss in this population. Tooth loss was associated with older age, destructive periodontitis, higher dental caries experience and previous dental visit.

In the present population, 80% of subjects had lost one tooth or more, and the mean tooth loss was 5.09. Mean tooth loss was higher in Indians living in the southern part of the Xingu Brazilian National Park and in Guarani and Portiguara Indians than in the Kiriri population24., 25., 26. In contrast, Australian Indians have experienced lower numbers of missing teeth²⁰ Higher tooth loss has been reported among non-Indian Brazilians in comparison with the Kiriri Indians¹¹ Adults in the USA and Japan experience lower rates of tooth loss: adults in the USA have an average of 24.92 remaining teeth²⁷ whereas Japanese adults have 23.5 remaining teeth, on average²⁸ One study of a Japanese population showed a mean tooth loss of 0.6, and 31.4% of the subjects had lost one tooth or more²⁹

Factors associated with the loss of one tooth or more were older age (≥35 years), severe periodontitis, higher dental caries and previous dental visit. Higher number of teeth lost has also been associated with older age in other populations15., 17., 25., 27., 30., 31., 32.. The association with age ≥35 years may be a result of increasing periodontal attachment loss over time³³, as attachment loss is higher in older people16., 34.. In agreement with the present study, periodontitis and attachment loss have been considered important reasons for extraction of teeth30., 35., 36.. Higher dental caries experience was associated with loss of one tooth or more, in agreement with previous studies30., 37.. This result suggests that dental caries and periodontitis are the main cause of tooth loss in Kiriri Indians. Previous dental visit was associated with greater tooth loss in this population. This association has also been described in other populations5., 32.. Not going to a dentist may be associated with retention of a tooth that should be extracted³⁸. Lower frequency of dental visits was significantly associated with incident tooth loss in previous studies11., 36.. The reasons for tooth extraction involve physiological and socio-economic factors, personal attitudes about the value of retaining natural teeth³⁹ and dental care philosophy. Dental services in the present population are mainly focused on tooth extraction, which explains the association between tooth loss and previous dental visit. Furthermore, this association may be a result of inequalities in oral-health practices⁴⁰

Smoking habit was not associated with tooth loss. This result contrasts with previous studies27., 29., 30., 31., 32., 36., 39., 41., 42. Although 50% of the Kiriri population were smokers, the mean number of cigarettes smoked per day was 3.7¹⁶ which is considered light smoking⁴³ and might explain the contrasting results.

Education was not associated with tooth loss after adjustment for covariates. This was probably because of the generally low educational level of the population. For comparison, tooth loss was significantly associated with low education level among non-Indian Brazilian, Japanese and US subjects11., 27., 43., 44..

In the present population, health care is still insufficient for preventing caries and periodontal disease. Access to dental services is mainly focused on tooth extraction, as observed by the high number of missing and decayed teeth. Reducing the number of missing teeth among adult individuals is a WHO global oral health goal for 2020⁴⁵. Therefore, this research is important, as the results can be used to mobilise and engage public resources to bring oral health-care services to Kiriri Indians, to prevent tooth loss and improve oral health, according to their needs.

In conclusion, the prevalence and extent of tooth loss were high among Kiriri Indians. Older age, periodontitis, higher dental caries and previous dental visit were associated with tooth loss. This study confirms the findings of previous studies on tooth loss in other populations.

Acknowledgements

The authors want to thank the Kiriri Indian Community for the opportunity to develop this study, the Special Indigenous Health District (National Health Foundation, Ministry of Health, Brazil) for political and logistical support during study development and the National Council of Technological and Scientific Development (Ministry of Science and Technology, Brasília, Brazil, process # 308475/2009-7 and 477377/2010-6).

Conflict of interest

There is no conflict of interest related to this study.